Skip to main content

Full text of "On the development and transmission of power from central stations;"

See other formats

This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project 
to make the world's books discoverable online. 

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject 
to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books 
are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover. 

Marks, notations and other marginalia present in the original volume will appear in this file - a reminder of this book's long journey from the 
publisher to a library and finally to you. 

Usage guidelines 

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the 
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing this resource, we have taken steps to 
prevent abuse by commercial parties, including placing technical restrictions on automated querying. 

We also ask that you: 

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for 
personal, non-commercial purposes. 

+ Refrain from automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine 
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the 
use of public domain materials for these purposes and may be able to help. 

+ Maintain attribution The Google "watermark" you see on each file is essential for informing people about this project and helping them find 
additional materials through Google Book Search. Please do not remove it. 

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just 
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other 
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of 
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner 
anywhere in the world. Copyright infringement liability can be quite severe. 

About Google Book Search 

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers 
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web 

at |http : //books . google . com/ 


University of California. 



University of California. 


Dowat& Xecture0 



Edition, Beriaed and BaUrged. Fabt I. Oeneral Principles, 
Fastenings, and TransmissiTe Machlaeiy. With S04 Diacnuns and 
lUustrations. Grown Sro. 6«. 

I Edition, Revised and Enlarged. Fart 11. Chiefly on Engine Details. 

I With 174 Diagrams and IlJuatFations. Crown 8vo. U. 6i. 

I A Text-book for the Engineering Laboratory, and a Collection of the 

Resnlts of Experiment. 8to. Five Plates, 140 Woodoats. 91«. 

28 Plates. Fcp. folio. 4«. M., in case. 

' Londoa : LONQMANS, GREEN, ft CO. 

to facilitate Practical Calculations, and for Solring Arithmetical 
Calculations in Class. U. 6cl. 

London : E. ft F. K. 8P0N. 







Ube f)oward Xectures 


BY , 


B^a, M. i2caT. CIVIL msQtJmB 


rBonsflOR or enoinkssinq at the guilds central technical collbob 




All right* rtttrvtd 


The present Treatise is based on the Course of Lectures which 
the Council of the Society of Arts requested the Author to 
deliver in January and February 1893. In republishing them 
under less stringent limitations of time and space many gaps 
have been filled up and some questions have been discussed more 
fully. The importance of the problem of distributing power to 
many consumers can hardly be overrated. In dealing with it the 
question of cost cannot be put on one side. The financial condi- 
tions are governing conditions, and must be considered together 
with the mechanical conditions. An attempt has been made in 
the present treatise to treat the subject as a whole. Hence the 
causes of waste in generating power have been discussed as well 
as the losses in distribution. The subject is so wide and 
touches so many departments of engineering that it is too 
much to hope that all the questions involved have been examined 
with sufficiently adequate knowledge. 

But much care has been taken to indicate what is essential 
in the consideration of schemes of power distribution, whatever 
the source irom which the power is obtained and whatever the 
method of transmission adopted. For the rest, practical ex- 
perience will gradually determine much that is at present 

Hat, 1894. 




I. The Conditions in which a System op Disteibution op 



n. FowEB Genebated by Steam Engines. Conditions of 

Economy and Waste 19 

in. The Cost of Steam Poweb 68 

IV. The Stobaob of Eneboy G7 

V. Wateb Poweb 80 

VI. Hydbaulic Motobs 95 

VII. Telodynamic Tbansmission 108 


IX. Tbansmission of Poweb by Comfbbssed Aib . . .163 

X. Calculation of a Compbbssbd Aib Tbansmission when 
the Subsidiaby Losses of Eneeoy abb taken into ac- 
count 211 

XI. Disteibution of Poweb by Steam 216 

XII. Disteibution of Gas fob Poweb Pueposss . . . 265 

XIII. Elbcteical Tbansmission of Poweb . . . «l . 263 

XIV. Examples of Poweb Tbansmission by Elbcteical Methods 286 

XV. The Utilisation of Niagaba Falls 297 

Index ^809 

»f THE 






The late Mr. Thomas Howard left a bequest to the Society of 
Arts, to provide for the delivery, periodically, of courses of 
lectures relating to the production and use of motive power. 

In carrying out the duty imposed by this trust, the Society 
did the author the honour of inviting him to give some lectures 
on the development of power at Central Stations, and its dis- 
tribution, either as motive power for driving factories and 
workshops, or as energy applied to other purposes. Energy is 
in these lectures to be considered as a commodity, which can be 
manufactured in a convenient form, and distributed and sold. 
The special problems to be dealt with are the conditions which 
favour the production of a convenient form of energy, on a large 
scale and in the most economical wfiy ; the means of conveying 
it to a distance and distributing it to consumers ; the arrange- 
ments for measuring the quantity delivered; and lastly, the 
relative advantages and disadvantages of a system in which 
energy is obtained on a large scale and distributed to many 
consumers, compared with a system in which each consumer 
produces the power he requires in his own locality and under 
his own supervision and responsibility. 

As various sources of energy are available, and as there are 
several methods of distributing energy or means of readily 
obtaining energy, the inquiry is a wide one. It may be limited 



at once to those cases in which the final form of energy required 
is mechanical energy. Pressure water is a convenient source of 
mechanical energy, so that a water-supply system with its 
pumping station and distributing mains is from one point of 
view a central station from which energy is distributed. But 
water-pressure systems need only be considered in these lectures 
so fisir as they are actually used for distributing motive power. 
An electric lighting station is a central station distributing 
energy to many consumers. But it will only come within the 
scope of these lectures to consider electrical systems, in which 
the whole or part of the electricity distributed is used for 
motive power purposes. The addition of motive power supply 
to the primary object of water supply or light supply may be of 
importance, and it is only so far as a combination of purposes 
in such systems facilitates or cheapens the supply of motive 
power that they will be considered. 

That the author has more knowledge of some of the older 
and more mechanical methods of power distribution than of the 
newer methods of electrical distribution, is no doubt a dis- 
advantage. But the distribution of electricity both for lighting 
and for power purposes has already been fully discussed, both in 
special treatises and especially in the lectures of Prof. George 
Forbes and Mr. Gisbert Kapp. On the other hand, there may 
be some partly compensating advantage in approaching the 
subject for once with the bias of an engineer rather than that 
of an electrician. Granting that electrical distribution will 
play an important part before long in the development of 
systems of power distribution, there is a popular tendency at the 
moment to regard too exclusively electrical methods, and to 
overlook other means of ppwer distribution which have been 
usefully applied in the past, and will, in suitable conditions, be 
still employed in the future. 

Unscientific people who see the electric lamp are perhaps a 
little apt to attribute to electricity too exclusive a share in the 
production of the light. They forget the coal, the boilers, and 
the steam engines, which are as necessary as the dynamo in ob- 
taining the result. Perhaps, with less unscientific people, the 
striking success achieved in transmitting energy electrically to 
great distances, and retransforming it when necessary into 
mechanical energy, has obscured the fact that there are other 


methods of power transmission more convenient and less costly 
in particalar cases. 

Two points should be clearly kept in mind. First, that, 
€is to the production of energy in an available form, we are just 
where we were before modem electrical discoveries were made. 
The most the electrician can do is to provide a new mechanism 
of distribution. The case would be different if, by means of 
primary batteries employing new materials for chemical action, 
or by thermo-electric batteries in which heat is directly trans- 
formed into electricity, the methods of producing available 
energy were changed. Such things are theoretically possible. 
Practically they are not yet available. At present the energy to 
be distributed must be developed by a steam-engine or water- 
wheel, and the dynamo, cable and electric motor merely replace 
the shafting, gearing and belting, or other mechanism of trans- 
mission, performing the same functions, more cheaply and 
effectively in certain cases it is true. 

The second point is that every method of transmission will 
be found to have some characteristic advantages fitting it 
specially for particular cases. It may be conceded to the 
electrician that the special advantages of electrical transmission 
are strikingly apparent, where power must be conveyed to very 
great distances. But such cases are likely to be comparatively 
rare. The remarkable mechanical and scientific success of the 
Frankfort-Laufien experiment, in which 300 h.p. was conveyed 
108 miles with a loss in transmission of only 25 per cent., 
has a little misled merely popular observers. The fact must 
be borne in mind that the cost of the power, when it reached 
Frankfort, was five times as great as that of an equal amount 
of power produced directly in Frankfort by a steam-engine. 
It can only be, where some exceptionally cheap source of power 
at a distance can be put in competition with dearer power 
near at hand, by means of electrical transmission, that long- 
distance transmission is likely to be applied. For trans- 
missions to moderate distances there is a choice of several 
means of transmission, and electrical distribution has not 
in such cases and up to the present time established any 
universal superiority. 

The cost of production of mechanical energy from natural 
sources has to be considered in these lectures for several reasons. 

B 2 


If systems of producing energy in a wholesale way at central 
stations and for its distiibation in parcels to different consumers 
should be adopted, it will chiefly be because energy can be so 
obtained from sources so much cheaper, or at so much less cost of 
production, that it can be distributed and sold at less cost than 
that at which it can be produced by individual consumers. It 
is necessary, therefore, to consider in what ways central station 
working renders available new sources of energy or is favourable 
to economy — what, in fact, is the relative cost of energy obtained 
in central stations on a large scale and in isolated workshops 
on a smaller scale. The methods and cost of distribution form 
a second branch of the inquiry. 

Oeneral Statement of tJie Advantages of Central Staiions for 
Producing Ene^-gy. — Some preliminary general considerations 
may be stated in favour of systems of producing energy at 
central stations and distributing it. 

1. In generating power by steam-engines and boilers there 
is obviously some advantage in cost of machinery, in economy 
of fuel, and in cost of superintendence due to concentration of 
the engines and boilers in a single station. The central station 
can be placed where coal can be delivered cheaply, and arrange- 
ments can be adopted to secure economy which would be too 
costly or too complicated in small plants. 

2. In the case of water power, it often happens that it is only 
possible to deal with a natural waterfall either by a combination 
of consumers, or by an association acting in the interest of many 
consumers, for the construction of the costly permanent works 

3. The locality where the power can be generated may be 
fixed by one set of conditions, that where the power must be 
used by another set of conditions. Often it is a question of 
adopting a cheaper source of power at a distance or a dearer 
source near at hand. Thus, in mining and tunnelling operations, 
cheap steam power at the surface may be distributed to replace 
much more expensive hand labour in the workings. That is 
essentially a case of power distribution from a central station. 
Mr. Thwaite has recently proposed to erect large gas motor 
stations at collieries and to transmit the power electrically to 
the nearest in4ustrial centres, where the power could be used 
in manuPacturing operations.^ Whether such a scheme is 

» The Engineer, December 2, 1802. 


advantageous or not is mainly a question of finance. The 
question to be solved is, whether there is so much economy in 
producing energy at the colliery that it can be delivered in the 
form required at a distant point, at a cost less than that at which 
it could be produced on the spot where it is used ; for instance, 
from coal brought from the colliery by railway. In utilising 
water power, a similar problem often arises. The water power 
can only be utilised at a point distant from the manufactory 
where power is required. Then it has to be considered whether 
the transmitted water power costs more or less than power 
obtained at the manufactory from coal. 

4. Another reason for central-station working, in large towns, 
is the great inconvenience of generating power in the small 
quantity and at the irregular times at which it is wanted for 
many purposes. For good or for ill, population gathers into 
huge communities, in which there is a complex development of 
social and industrial life. In such communities there is a 
constantly increasing need of meclianical power. In addition 
to manufacturing operations, demands for power arise for transit, 
for handling goods, for passenger lifts, for water supply, and for 
sanitation. At first these are met by the erection of scattered 
motors. But this sporadic production of power in small quantities 
is quite certainly in many instances extravagantly costly and 
inconvenient. There is a great probability that power dis- 
tributed from a central station in towns would be so convenient 
as to be preferable to power produced locally, even at a some- 
what greater cost. 

Just as it has become necessaiy to supersede private systems 
of water supply by a municipal supply, just as it has proved 
convenient to distribute coal gas and necessary to establish a 
general system of sewerage, so it will probably be found con- 
venient and even necessary to provide, in towns of a certain 
importance, some means of obtaining mechanical power in any 
desired quantity, and at a price proportional to the amount of 
power used. It is socialism in the field of mechanical engineer- 

For the single purpose of working lifts and hoisting 
machinery, it has already proved remunerative to extend 
through tiie streets of London a system of hydraulic mains nearly 
sixty miles in length. The extremely rapid extension of the 


system is worthy of note. In July 1884, there were only 96 
consumers taking power from the London Hydraulic Company's 
mains. In 1888, there were 720 consumers renting power. In 
1892, there were 1,676 consumers renting power, and the use of 
the system is now extending more rapidly than at any previous 
period. The quantity of water distributed has increased from 
317,000 gallons per week in 1884 to 6,600,000 gallons per 
week in 1892. In no instance has the use of hydraulic power 
when once adopted been abandoned in favour of any other 
system of working hoisting machinery. 

There can be no doubt that when (a) power can be obtained 
with little trouble, in a form involving no great amount of 
superintendence in working, and (6) the cost is proportional to 
the amount of power used, then a demand for the power is- 
readily created. 

Perhaps a more striking instance of the growth of a demand 
for power is furnished by the town of Geneva. No casual ob- 
server would have selected Geneva, with its population of 50,000, 
as a likely centre for a great system of power distribution. Yet 
the works at Geneva, which will be more fully described later, 
are perhaps the most important example of power distribution 
hitherto carried out. In 1871, Colonel Turrettini obtained 
permission from the Municipal Council to place a water-pressure 
engine on the existing low-pressure town mains, for driving the 
factory of the Genevan Society for the Manufacture of Scientific 
Instruments. In case of this method of obtaining power proving 
satisfactory, he obtained the right to instal similar motors in 
other parts of the town. The plan proved so convenient that 
nine years afterwards, in 1880, there were 111 water motors 
driven from the low-pressure town mains, using 34,000,000 
cubic feet of water annually, and paying to the Municipality 
2,000Z. a year. The cost of the power was not small. It was 
charged for at 3d. to 4cZ. per h.p. hour, which is equivalent to a 
rate of 36J. to 48 J. per h.p. year for motors working continuously 
for 3,000 hours in the year. Since that time a new high- 
pressure service has been established, the water being pumped 
by turbines in the Rhone. On the high-pressure service the 
cost of the power is less. It is charged for at about 0'7d!. per 
h.p. hour, equivalent to 8Z. per h.p. year of 3,000 working hours. 
In 1889, the annual income from power-water sold on the low- 


pressare system was 2,085Z., and on the high-pressure system 
4,500Z, At that time the receipts for power-water were in- 
creasing at the rate of 880Z. per annum. In 1889, the motive 
power distributed on the high-pressure system amounted to 
1,500,000 h.p. hours, annually, there being 79 motors, aggregat- 
ing 1,279 horse power. 

This illustrates sufficiently the growth of the use of motive 
power distributed in a convenient form. The power used in 
pumping the ordinary water-supply for municipal and domestic 
purposes is not included. It will be seen later on that the 
works, taken as a whole, are very large and important. 

The location of the windmill on the hill and the water-mill 
by the stream indicates how conditions of human labour have 
been determined by the need of mechanical energy. The earlier 
cotton mills were all placed where water power was available, 
although this had the disadvantage of taking them away from 
the places where skilled workmen were found and from the 
markets for manufactured goods. In an interesting pamphlet 
on the ' Eise of the Cotton Trade,' by John Kennedy, of Ardwick 
Hall, written in 1815, it is stated that, for some time after 
Arkwright's first mill was erected at Cromford, all the principal 
miUs were built near river falls, no other power than water 
power having been found practically useful . After the invention 
of the steam-engine, manufacturing industries gathered round 
the coal-fields. About 1790, says Mr. Kennedy, *Mr. Watt's 
steam-engine began to be understood and waterfalls became of 
less value. Instead of carrying the people to the power, it was 
found preferable to place the power amongst the people.' The 
tendency of the conditions created by the introduction of steam 
power was to concentrate the industrial population into large 
communities and to confine manufacturing operations to large 
factories. The economy of producing power on a large scde 
and the difficulty which then existed of distributing it to 
any great distance favoured the growth of the factory system. 
Facilities for distributing power to considerable distances are of 
more modern origin, and may partially reverse the tendency to 
concentration. Further, facility in conveying power will permit 
the utilisation of some sources of power hitherto not available. 
Waterfalls are most commonly in positions inconvenient for 
industrial operations. But if tlie power generated at the fall 


can be transmitt'ed to localities where it can be conyeniently 
used, the availability of water power is greatly increased. 
Mountain districts where water power is abundant may come to 
have an advantage over districts near coal-fields. 


In these lectures motive power is treated as a commodity, 
prodncible, distributable, saleable. The first question is as to 
the sources from which it can be obtained. 

Wind power has been used for driving ships and mills, and 
now and then it is alleged that, as a source of power, wind action 
has been too much neglected. But its intermittence restricts 
its use to work which can be intermittent also. The compara- 
tively short periods in which the wind pressure is a considerable 
force make it uneconomical to attempt to do more than to utilise 
very moderate winds. On the other hand, the occasional great 
intensity of wind action during short periods involves the 
necessity that structures exposed to its action should be of ex- 
cessive strength and costliness. 

Tidal action might, no doubt, afibrd an enormous amount 
of mechanical energy. But, up to the present time, it has been 
found that the cost of embankments and machinery for utilising 
tidal action is so great as to prohibit its employment. The 
direct action of the sun's heat could be employed, but here 
again the cost of utilisation exceeds the value of the power ob- 
tained. Considered practically and commercially, there are only 
three sources of mechanical energy of industrial importance : — 
(1) the muscular energy of animals ; (2) the work of water 
falling fix>m a higher to a lower level and automatically restored 
to the higher level by the sun ; (3) the heat obtained by the 
combustion of fuels, transformed into mechanical energy by beat 

The muscular energy of animals need not be considered in 
these lectures. It will be convenient to consider energy 
derived from the combustion of fuels before considering water 
power. By far the most important source of mechanical energy 
for industrial purposes is solid fuel burnt in the furnace of steam- 
boilers. The heat produced is used to generate steam, from 
which mechanical energy is obtained in a steam-engine. There 



4ure, however, disadvantages in obtaining energy from coal. 
First of all, at least one-fifth of tixe heat generated in a boiler 
furnace by combustion of the fuel escapes by radiation and in 
the chimney. Next, there is the fundamental disadvantage that, 
under possible temperature conditions, not more than three- 
•eighths of the heat given to the steam (that is, three-tenths of 
the heat of combustion in the furnace) can be transformed in the 
engine into mechanical work, the remainder being necessarily 
rejected in the condenser. Of this 30 per cent, of the original 
heat energy there are further large fractions wasted, in conse- 
quence of practical imperfections of the steam engine. Cheap 
as coal is, the cost of the energy obtained from it is multiplied 
greatly by unavoidable imperfections of the processes of trans- 
formation. Further, the attendance required in operating boilers 
and engines, the fire and explosion risk, the nuisance of smoke 
and difficulty of getting rid of ashes, are all drawbacks attending 
the use of steam power. 

Solid F'ltel. — By far the most important source of 
mechanical energy is the combustion of various natural solid 
fuels, chiefly the different descriptions of coal. Coal is ob- 
tainable in very many localities, and can he transported even 
great distances at comparatively small cost. The solid fuels 
owe their thermal value almost entirely to the carbon and 
hydrogen they contain. It has been ascertained that the heat 
produced by the combustion of these constituents of fuel is as 
follows : — 

Combustible constituentfl 

Product of combustion 

Tlicrmal Units 

(Pound (legreea, 


combustion of 

one pound 

Weight in 

pounds of 

products of 



Carbonic oxide 

. I Carbonic anhydride 
. Carbonic oxide 

Carbonic anhydride 

Water . 

Steam . 











The natural solid fuels may for practical purposes be con- 
sidered to be composed of carbon, hydrogen, and a portion of 
incombnstible impurity. In a good furnace the air supply must 
be sufficient to ensure perfect combustion of the carbon into 
carbonic anhydride, or, as the table shows, there will be a waste. 



Usaally the products of combastion must escape at Bach a tem- 
perature that the product of combustion of the hydrogen escapes 
as steam. Hence the amoant of heat obtainable from a pound of 
coal containing c lb. of carbon and h lb. of hydrogen is at most 

Q=: 14540 c + 52830/t 
=14540(c + 3-63A)Th. U. 

An exact determination of the calorific value of a coal can only 
be made by direct calorimetric test.^ But the formula gives 
very approximate results for ordinary solid fuels with one small 
correction. In the incombustible part of the coal as actually 
used is included a portion of hygroscopic water, which has to be- 
evaporated at the expense of the heat produced. Each pound 
of hygroscopic water absorbs in evaporation about 1000 Th. U* 
If a pound of coal contains w lbs. of water, then its thermal 
value used in a furnace from which the water escapes as steam, 
and neglecting at present the radiation and chimney waste, is, 

Q=14540(c + 3-63A)-1000ir Th. U. 

The following table gives the composition and thermal value of 
some ordinary fuels. It has been assumed that in ordinary 
solid fuels as burned there is at least 3 per cent, of moisture : — 



I Evaporation 

I Tliermal In lbs. of 

W.t.r|A.h.*«.; -;^P- ipJSTofJLl 
I I pouud of fuel from and at 

A , IT 2ir 


Wood (air dried) 
Brown ooal . 
Coke . 

»» • • • 
Welsh steam coal 

Staffordshire coal 
Anthracite . 
Bituminoas coal 
Poor small coal . 
Illnminating gas 
Petroleum . 
Gunpowder . 

. -85 

i -85 

I •ss 

' •SI 

I -70 



•06 i -08 
•20 ' -36 































* Direct calorimetric tests are difficult, and many such tests are unsatis- 
factory. The best tests seem to show that the true calorific value of a fuel 
cannot be deduced from its analysis. The possible error in calculating the^ 
calorific value from analysis may amount to i: 6 per cent. 


It will be seen that the ordinary coals do not differ greatly 
in heating value. The last column contains the evaporative 
value (at 966 Th. U. per lb. evaporated), supposing all the heat 
produced to be usefully expended in causing evaporation. 

Liquid Fuel. — Oils which are mixtures of various hydro- 
carbons, and which are known as parafiSn, shale oil, petroleum, 
or kerosene, may be obtained by the distillation of certain shales, 
and occur naturally in many localities. The production of oil 
for lighting and lubrication by distillation from shale, established 
first in Scotland, had assumed considerable importance when the 
oil wells of Pennsylvania were discovered in 1859. Since then 
the trade in these mineral oils has assumed enormous pro- 
portions, and new localities where such oils can be obtained 
have been discovered in many parts of the world. The table 
above shows that petroleum fuel has a high heating value. A 
pound of petroleum, if the heat were fully used, would evaporate 
35 per cent, more water than a pound of Welsh coal. In 
South-East Russia petroleum refuse is largely used as fuel for 
locomotive and stationary boilers. Such fuel appears to have an 
evaporative power 25 per cent, greater than that of good coal. 
Mr. Holden has used some crude hydrocarbon oils in locomotives 
on the Great Eastern Railway. More lately refined petroleum 
oils have been used in internal combustion engines of the same 
type as the gas-engine, with very satisfactory results. The 
consumption of the oil in such engines is only 0*85 lb. per indi- 
cated h.p., or 1 lb. per brake h.p. per hour. 

The refined oils which come to this country cost at least 
S^d. pep gallon, or ^d, a pound, or about 4Z. 68. per ton, a very 
high price compared with that of solid fuel. They cannot there- 
fore be used here under boilers in producing energy on a large 
scale, notwithstanding their high calorific value. Used in in- 
ternal furnace engines, which have a high thermal efiiciency, they 
produce energy at a cost somewhat lower than that at which it 
can be produced from lighting gas, at ordinary prices, and lower 
than that at w^ich it can be produced by small non-condensing 
steam-engines, which are thermally of poor efiiciency. The 
crude oils cannot be transported by ship and railway safely, 
and are not therefore to be obtained, in this country, at any 
rate in large quantities. 

It may, however, be noted that there is no difficulty in 


transpoi*ting the crude oils by pumping through pipes, and 
crude oil is now transported in that way considerable distances 
in the United States, and is used in boiler furnaces on a large 
scale, and apparently with economical advantage. During the 
present year the whole of the boilers at the World's Fair at 
Chicago, developing about 20,000 h.p., were worked with crude 
petroleum, pumped direct from the wells to a store tank near the 
boilers. From this the oil flowed by gravitation to the nozzles 
at the boilers, and was sprayed by a steam-jet into the furnaces. 
The facility of control and the absence of dirt and smoke were 
very great advantages. 

ITie same oil was supplied for working the boilers at the 
Hyde Park Water Works Pumping Station at Chicago. The 
author was informed that the cost of the oil was 60 cents per 
barrel (42 gallons). This is equivalent to about IGs, per ton. 
It was stated that, at the price at which it was supplied, the oil 
was a more economical fuel for raising steam than coal at 129. 
a ton. 

Gaseoiis IhieL — Many of the disadvantages of solid fuel 
are diminished by using the coal to produce gas, and then 
generating power by burning the gas in internal combustion or 
gas-engines. Gas can be transported with great convenience 
in pipes, and ga&-engines work with less attendance and higher 
thermal efficiency than steam-engines. In transforming heat 
into work, small gas-engines are enoi^mously more efficient than 
small non-condensing steam-engines and boilers. On the other 
hand, ordinary lighting gas, taxed as it is with costs of dis- 
tribution due to its ordinary application for lighting purposes, 
is more expensive for a given calorific value than coal. The 
cost of ordinary lighting gas is increased both by the need of a 
large generating plant, to meet the excessive fluctuation of 
demand for lighting, and by the large distributing charges 
involved in supplying a very great number of small consumers. 
If gas were made specially for heating and power purposes, 
either coal gas of low luminous power, or water gas, or producer 
gas, it could probably be distributed to power users at less than 
half the present price of coal gas. Used in gas-engines, it 
would then compete on nearly equal terms as regards cost with 
solid coal, and would have many subordinate advantages. 

M. Aira6 Witz has shown by direct experiment that a gas- 


engine, worked with Dowson gas, will give an effective horse 
power at a total cost, including all charges for fuel, interest, and 
depreciation, not greater than that at which an effective horse 
power can be obtained by a good boiler and good compound 
steam-engine. It is impossible to predict how far gas-engines 
will replace steam-engines, but at present they have two dis- 
advantages. They are more restricted in size than steam- 
engines, and work with less efficiency at light loads. 

Production of Power by Burning Tovm Refuse. — There is 
another source of heat energy — another fuel — which is of 
some importance in counection with the question of power 
distribution. In addition to ordinary sewage, which is disposed 
of in well-known ways, there is in large towns a quantity of 
ash-bin refuse and trade refuse, which can only be got rid of 
effectually and innocuously by burning. This refuse may amount 
to from 100 to 400 tons per 100 inhabitants per annum. In 
a steadily increasing number of towns, such refuse is now 
consumed in furnaces commonly termed destructors. There are 
thirteen towns in Lancashire and seven towns in Yorkshire 
which bum refiise in destructor furnaces, and there are similar 
furnaces in towns in other counties. Altogether there are about 
fifty-five towns in Great Britain, in which destructors are used. 
The refuse is reduced by burning to about one-third of its 
weight and one-fourth of its bulk. The organic matters and 
cinders in the refuse form the fuel necessary for combustion. 
The residue after combastion consists of clinkers and sharp 
ashes, which have a small saleable value. 

The refuse burned in destructors'varies greatly in composition 
in different localities. An analysis of London ash-bin refuse 
gave the following components : * — 

Per cent. 
Breeze, cinders, and ashes . . .04 

Fine dust 19 

Paper, straw, and organic matters . .12 
Bottles, bones, tin, crockery, &c. . . 5 

Even when the heat generated is wasted, it appears that 
burning in destructor furnaces is the least costly method of 
refuse disposal. But the amount of heat generated is not 

» Beport on Dust Deftructort by the JUediral Officer and Engineer of the- 
Landon County Council^ 1893. 


inconsiderable. If the Iieat coald be utilised as satisfactorily 
as that of coal in generating steam, the refuse would have a 
calorific value equal to about one-fifth that of an equal weight 
of coal. To any extent to which the heat can practically be 
utilised there is an additional advantage in this mode of dis- 
posal. The heat of destructor furnaces is energy which, like 
water power, costs nothing, except for the additional machinery 
which must be added to the destructor furnace for the purpose 
of utilising it. 

Various forms of destructor furnace have been tried. Five 
are described in a paper by Mr. Watson, of Leeds, read before 
the British Association in 1892. All these ai*e to some extent 
modifications of the Fryer Destructor, introduced in 1876. The 
destructors first built were comparatively low temperature 
furnaces. Temperatures taken by a Siemens pyrometer, in a 
destructor in Whitechapel, ranged from 180° to 1,000° F. The 
average of eight cells was 490° F.* At Ealing, the temperature 
in the flues averaged 631° F. With this low temperature the 
fames are ofiensive, and hence it has been necessary in many 
cases to have a second furnace, termed a fume cremator, in which 
coal or breeze is burned, and through which the products of 
combustion from the destructor pass on their way to the chimney. 
In more recently constructed furnaces, the aim is to obtain a 
higher temperature in the destructor furnace itself, by modifying 
the form of the furnace and especially by using a forced draught. 
In many recent destructors, the temperature in the furnace is 
from 1,500° to 2,000°, and this high temperature required for 
the inofiensive disposal of the refuse is, as will be seen presently, 
favourable to the utilisation of the heat generated. 

Frijer's Destructor. — This has up to the present been more 
extensively adopted than any other. In its original form, it 
was a comparatively low temperature furnace. It consists of a 
block of * cells ' or furnace chambers internally about nine feet 
long by five feet wide. The cells are placed in two rows back 
to back, and the products of combustion pass away by a common 
•central flue. The top of each cell is arched and the floor forms 
a grate, usually sloping at one in three towards the front, where 
there is a door for removing clinker. The refuse is charged at 

» lleport on the Destruction, of Tavm, Htfiue, By Thomas Codrington. 1888. 
Also Watson, SefMe IHspoialt British Association, 1892. 


the back and, as it burns, is dragged forw^ard along the sloping 
grate. Each cell will bum from eight to ten tons of refuse per 

At Southampton a destructor with six cells of this type 
has been erected,^ and arrangements are made for utilising part 
of the waste heat. Each cell will bum from eight to eleven tons 
of refuse per day. The products of combustion pass through a 
multitubular boiler, and thence to a chimney, 160 feet in height 
and 6 feet in diameter at top. There is a bye-pass through 
which the gases can be taken to the chimney without passing 
through the boiler. The steam generated is used to drive 
engines of 31^ i.h.p., which compress air for working three 
Shone ejectors. There are also engines for preparing fodder 
and for electric lighting. 

The garbage and ash-bin refuse is collected in covered iron 
tumbler carts of two cubic yards capacity. These are taken up 
an inclined roadway to the charging platform of the destructors. 
The residue after combustion consists of about 20 per cent., by 
weight, of the refuse burned of hard clinkers and sharp ashes. 
The former is used in road-making and the latter for mortar. 
The initial cost of the destructor, including engine-house, 
inclined roadway, chimney, boiler, and ironwork, was 3,723Z. 
The annual cost of burning, two men being employed by day 
and two at night, is 251Z. The quantity of refuse burned daily 
is about fifty tons. The minimum quantity burned in a day is 
twenty-five tons, which is sufiicient to keep up steam in the 
boiler. This gives about seventy-five pounds of refuse per 
i.h.p. hour actually utilised. 

HortfalPs Destructor. — This may be distinguished as a com- 
paratively high temperature destructor, the gases escaping at a 
temperature of 2,000° F. The cells or furnace chambers are 
five feet wide. The refuse is introduced through a large hopper 
at the upper end. The gases escape through openings in the 
furnace crown, and are thence led downwards to the main fine. 
The grate bars are of the rocking type, with a moderate amount 
of motion. The grate surface is 28 square feet per cell. The 
clinker is removed at the front. Nine tons has been burned per 
cell per day. The chief peculiarity of the Horsfall furnace is a 
closed ashpit, with a steam jet producing an air pressure of 
» Proe. Inst. Mech. En^^ineert, 1892. 



about half an inch of water column. It is due to this that a more 
active combustion is secured, with a thicker fire and with a less- 
air supply. Consequently, the products of combustion escape at 
a higher temperature, which is better for merely destructor 
purposes, and at the same time makes the escaping heat more 
available for utilisation. 

Modified Fryer Dedructw at Leeds. — Fig. 1 shows a 
destructor erected by Mr. Hewson at Leeds, which combines to 
some extent the features of the Fryer and Horsfall furnaces. 



FlO. 1. 

The refuse is introduced at the back, the outlet flue opening is 
at the front of the furnace, and the gases are led back between 
the cells to the central flue between the two rows of cells. 
There is a forced draught obtained by a steam jet in a closed 

Galoiific Value of Ash-bin Fefuse. — In many cases where part 
of the heat from destructor furnaces has been utilised the arrange- 
ments have been imperfect, and only about four to six horses 
power have been obtained per cell ; that is, about a hundred- 
weight of refuse burned per hour yields a horse power. This 


would probably correspond to about 4 lbs. of refuse to evaporate 
a pound of water — a very low result. 

A calculation by Mr. Watson on data obtained in working a 
Horsfall furnace at Leeds, in 1888, led him to conclude that a 
pound of refuse would evaporate about 2 lbs. of water, if the 
loss in radiation and chimney waste did not exceed 20 per cent. 
of the heat produced. Looking at the composition of refuse it 
does not appear that this estimate is excessive, but the data 
given are not very satisfactory and better data are at present 

According to Mr. Keep,* it has been found at Birmingham 
that on the average 1 lb. of refuse will evaporate 1*79 lbs. of 
water ; and at Warrington that 1 lb. of refuse will evaporate 
1-47 lbs. of water. 

If the heat is utilised to produce steam, by taking the gases 
from the destructor through a boiler, only that part of the heat 
which corresponds to the difference of the temperatures at which 
the gases enter and leave the boiler is utilised. Suppose the 
air enters the destructor at 60**, is raised in temperature to 
1 ,500** in the destructor furnace, and after passing through the 
boiler is discharged into the chimney at 500**. Then the fraction 
of the heat generated utilised by the boiler is, — 

l,500_-.500 _ 1,000 
1,500-60 1,440 

or about five-sevenths. If the destructor works at 2,000**, the 
heat utilised in similar conditions would be, — 

2,000-500 ^ 1^00 
2,000-60 1,940 

or about three-fourths. If the heat is to be utilised it is im- 
portant to work the destructor at as high a temperature as 

Mr. Watson made some other experiments at Oldham. Six 
cells were used, burning 1^ tons of refuse per hour. The gases 
passed through a multitubular boiler 7 feet in diameter and 
12 feet long, and the feed was measured by a meter. The 
temperature of the gases was 2,019® before reaching the boiler 
and 900° after leaving it. Thus, only about two-thirds of the 

* UtiU^Uicn of Tomn'i Brfme. C. C. Keep. British Association, 1893. 


available temperatare range was utilised. In two trials the 
mean evaporation was found to be 2,780 lbs. per hour. Deducting 
1,500 lbs. of steam used for the steam jets there was a surplus 
evaporation of 1,283 lbs. per hour, or, say, 50 h.p. of energy 
from six cells, burning together 1^ tons per hour. If a boiler 
with larger heating surface had been used, possibly an evaporation 
one-third greater would have been obtained, or say 3,700 lbs. 
per hour. Deducting 1,500 lbs. used in steam jets there would 
be a surplus or available steam supply of 2,200 lbs. per hour, or, 
say, 88 effective h.p. from six cells burning a total of 1^ tons of 
refuse per hour. This is about 40 lbs. of refuse to the effective 
h.p. hour. 

The chief difficulty in using the available energy of destructors 
for power purposes is this. The refuse must be burned at a 
nearly regular ra,te. But demands for power are fluctuating and 
intermittent. Presently, a method of heat storage is to be 
described which overcomes this difficulty. The adoption of such 
a method would render the utilisation of the waste heat of 
destructors much more practicable, and would give to this source 
of energy a much greater importance in connection with the 
problem of distributing power in towns.* Of course, refuse is a 
very poor fuel, and there is considerable expense in labour in 
burning it. It would not, therefore, be chosen as a fuel for 
raising steam. But it has to be burned for sanitary reasons. 
If any profit can be made by utilising the heat, that is a gain. 
The cost of the power obtained is merely the interest on the cost 
of the boilers and appliances which have to be added to the 
destructor, in order to utilise the heat which would otherwise be 

' It may be mentioned here that Prof. G. Forbes has proposed to use the 
heat from deetractors to work steam pumps lifting water to a reservoir on a 
hill. A store of water power woald thus be continaonsly accumulated, which 
could be used to drive hydraulic motors at times when motive power was re- 
quired. It will be seen later that a constant water power due to the flow of 
a river is thus xitillsed, by pumping to a reservoir, for intermittent work, in the 
systems at Zurich and Geneva. Prof. Forbes's proposal is the adaptation of 
the same method to a new case. 




Sir Frederick Bramwell, in an address to the Institution of 
Civil Engineers in 1885, indicated in a convenient phrase those 
'Conditions involving waste of fuel in the production of steam 
power, which are unavoidable when separate engines are used, 
but which can be diminished by central-station working. He 
said that we were ^ every day becoming more alive to the benefit, 
-where little power is required, or where considerable power is 
required intermittently, of deriving that power from a single 
«ource.' Small steam-engines are nearly always costly, un- 
economical, and inconvenient. Large steam-engines and boilers 
working with a varying and intermittent load are in conditions 
nnfavourable to economy. It is necessary to examine these 
cases in detail, and to trace the causes of waste. That is one 
step towards understanding in what circumstances central-station 
working is desirable. 

Evaporative Power of Boilers at Full Load. — The follow- 
ing Table contains a selection of the results of the most 
carefully made tests of boilers. The boilers may be assumed to 
have been worked at nearly the full load, except where trials 
were made with a varying rate of evaporation. For comparison 
of different boilers, working with difierent feed and steam 
temperatures, the evaporation per pound of fuel is reduced to 
the equivalent evaporation from and at 2 1 2° F. Where possible, 
the influence of difierent qualities of coal in different trials has 
been eliminated by reducing the evaporation to the equivalent 
evaporation by a pound of pure carbon. 

The first thing to notice in this Table is that the evaporation 
per pound of coal does not vary in different boilers so greatly as 
might perhaps be expected, from the great variation of con- 
struction and of the conditions of working in different experi- 
ments. Thus taking the column which gives the evaporation 

c 2 





-„., Steam 
J;^- preuure 


NO.; ^^e^"' 


burned per' 
tsq. ft. of 

nower P" ^- ^' 

grate per 

- - - 


by gauge 



1 Portable (loco, type) 

Kennedy k, 





2 CJomish . 




S Lancashire 

Ellington . 

179 1 83 



»t • • 

Donkin . 



6-76 ' 


11 • • 




19'82 , 


tt • • 






f* • • 






»» • • 












ff • • 

Donkin 5c 

[50] 50 



ff • • 


[60] 57 



n • • 


- 1 79 



»> • • 






Babcock . 

Longridge . 




♦» • • • 






»» • • • 

Percy StUl 

214 150 



Tubular . 


262 ' 99 

7-36 I 


>f • • • 

t» • 

290 99 

8-62 1 


»» • • • 

Carpenter . 

574 , 122 



Water Tube (Thorny- 

Kennedy . 





ft • • • 






t» • • • 



29-8 ' 


tt • • • 




66-8 ; 



Marine . 


[200] ' - 

17-27 ; 

! 25 

It • • • 


[300] - 



Portable Agricul- 


[26]: - 






»» • • 




Meteor . 

Kennedy . 







371 57 





980 . 81 





1,087 144 





645 165 



Ville de Douvres . 


2,977 106 

31-3 1 



per sq. ft. 

of total 



surface per 














Figures in square brackets are approximate 




Water eraponitcd 
from and at il^. 



value of oool 
per pound in 

per pound 
as us 

of ooal 

By feed 

tion by 
boiler onlv 
per pound 
of carbon 


By boiler 












Welsh Steam . 




Seaborne Bough 
















Mean of 6 trials 






Same boiler as 





Mean of 2 trials 






Same boiler as 

Slack Coal 






Welsh Steam . 





Perret grate 

>t >» • • 





Perret grate 
(same boiler) 












Wigan Slack . 






Wordey Burgy 





Mean of 3 trials 

Welfth Merthyr 











Same boiler 

>» • • 






Welsh Steam . 






» »» ♦ • 



•Same boiler 






n ti • * 




Hartley Newcastle . 






>* •> • 





Welsh Steam . 




Specially skH- 
fal firing 

»♦ • • 




If »♦ 

Scotch Bitmninoos . 





Hartley Newcastle . 





— . 

Midland . 






Welsh . 





♦Corrected for 
moisture in 

Walbottle Tyne 






Block Fuel . 





- 1 

figares, assumed from general knowledge. 


per pound of a fuel equivalent in heating value to a pound of 
pure carbon, the variation of evaporation for different boilers is 
not very great. Excluding a case or two where the boilers were 
certainly working badly, the lowest evaporation from and at 
212^ is 9^ lbs., and the highest 13 lbs. per pound of fuel. But 
in these boilers a square foot of heating surface produced from 
1^ to 9 lbs. of steam per honr. 

Next, it may be pointed out that in boilers the evaporation 
per pound of fuel improves as the total quantity of steam pro- 
duced diminishes, or, in other words, as the boiler load diminishes. 
Taking trials 15 to 18, which were made on the same boiler, or 
trials 19 and 20, also made on the same boiler, the results agree 
pretty closely with the expression 

E = 13-5 - 0-4*^, 

where E is the evaporation in lbs. per pound of fuel, and ir the 
steam produced in lbs. per hour, from each square foot of heat- 
ing surface. Such a rule cannot be pushed to extreme limits of 
working, but the general bearing of all the trials is that the 
efficiency of the boiler is greater as the quantity of steam pro- 
duced diminishes. The boiler thus to a certain extent balances 
the converse action of the engine, which is less efficient for light 

Table II. contains the results of carefully selected steam- 
engine trials. In these trials it may be assumed that everything 
was working at its best, and that the steam and coal consumption 
given is the smallest realisable with the given engine, for the 
conditions of pressure, speed, and power stated. In these trials, 
also, it may be assumed that the load was constant and ap- 
proximately the best for the given conditions. 

Broadly, the steam consumption and fuel consumption are less 
for large engines than for small engines ; less for quick than for 
slow engines ; and for suitable pressures, less for compound and 
triple than for simple engines. Two special groups of tests 
havQ been selected to show the economy due to jacketing, and 
the economy due to the use of superheated steam. 

In the most favourable trial conditions, as will be seen by 
the table, and with an economical and constant load, there is 
great variation in the amount of steam and coal required per 
indicated horse power hour. 




Table ILa.— Non-condensing Engines 

: 1 :i 





No.' Tj-pe of engine Authority | | ^ 

■ 1 iT 





' 1 Simjfle 

ft per 

lbs. per 
aq. in. 



1 Bmall Tower Spherical Unwin 






593 revs, per 



. 2 Semi-portable . „ 







3 Horizontal, coupled slide English 29-7 





Same engine, 

valve 1 


» t» i» 36 





varying load 


1 .. 61 





' 6 

Small doable-acting . ' Don kin (> 







Horiz. McLaren . . Unwin 3-8 






. . ., 6-3 

242 ! 60 


[.Same engine, 

9 1 W • • f» 





1 varying load 


>* • • ' tt 








Willans (central valve) Willans 





slow ! 


; „ 1 20 

224 , 112 




Willans (central valve) 



394 I 36 


>■ Same engine 


^ « )t >» , »» 





15 1 „ „ ., 34 






16 Beam .... Hirn . 78-3 






steam used 

17 Wheelock . . . | Hill . 140 



24 9 


IH Reynolds Coriiss . . i „ 187 





I'J Harris CorUss 









1 Armington 

Meunier 84 






2 Willans (central valve; 

Willans 10 





Not jacketed 

slow j 

3 .. », „ . 11 

123 ■ 103 



'• M 1» t» 







5, Willans (central valve) i „ 






fast , 

1 « n ,. „ i 36 












, THple 

: 1 Willans (central valve) Willans' 39 





Not jacketed 



The figures in brackets in the last column but one are estimated on the assumption 
that 9 lbs. of steam are produced per pound of coal. This would be for most cases about 
10-7 lbs. from and at 212° F. 

' Brake horse-power, and steam per brake horse-power per hour. 



Table ILb.— Condensuto £^'OI2?BB 
Statianarjf JSnginet 















TTpe of engine 

Willans, fast . 

Salzer (Trois Fontaines). 
Corliss . 

»f • • • 
Salzer . 

Reynolds Corliss . 
Harris Corliss 

Balzer (Angsburg) . 
Sulzer . 


Willans, slow 

Willans, fast 

Tandem Mill 

Receiver Mill . 

»» »» 

Salzer (Alo«t) 
balzer . 

Salzer (Florelfe) 

Sulzer (Van Hoegaarden) 

Salzer (Belgium) . 

Willans, slow 

Willans, fast . 

n n 

bulser (Augsburg) 

»» If 


Sulzer (Buda Pesth) 
Sulzer . 


Willans . 




Unwin . 
WUlans . 





291*5 90 

284 87 


I Vincotte 

j Soldini . 

j Sulzer . 

I Vincotte 

Willans . 







































lbs. per 
' sq. in. 

' 5 

ft, perl 

380 : 

382 ! 
380 1 
272 , 
606 . 
372 1 


; 196 

i 300 



, 399 


I 394 

' 397 



I 442 
I 487 
' 478 
! 39J5 
; 590 
, 600 
I 493 

I 13-90 
' 13-35 

67 44 

231 170 

21-3 = 120 

29-5 170 

601 145 

302 16-99 

302 I 12-86 

384 13-39 

379 1302 

607 12-82 

596 12-45 

460 12-2 

' 516 11-85 

, — 11-70 [1-3] 












I Jacketed 


5 expan- 

> 10 exf an- 
i sions 
, 15 to 20 
>' ezpansioDs 

)^ Not 
) jacketed 





Tabls II.B.— Condbnbinq Engines.— C^^n«0<f 
Marine Engine* 

Type of eiigine 


A'ille de Donvres 


1 Tart4ir 

2 3Ieteor 

3 i lona 


1 ^2 a g.1 

Authority 84 || || g| 

Kennedy 1,979 
I 371 

llM.per I 

sq. in. ft. per 

gHUge mill. I IbB. 

81 I 520 ' 21-73 

67 306 , 21 17 

;2,977 , 106 442 i 20*77 

I ; I 

Kennedy 1.087 144 , 490 , 19-83 

1 1,994 145 574 ; 14*98 

645 1 165 397 1335 






I Jacketed 
H. P.Jackd. 

Pumping Engines 

No. Type of engine 

1 1 












' Simple 

lbs. per 
sq. iu. 

ft. pet 



1 Beam pumping 
1 2 „ „ . . 









} Jacketed 

1 Compound 

1 Tandem pnmping . 







Not jacketed 

2 , Worthington* 

Mair and 






3 1 .. ... 

4 ' Beceiver 

Mair . 








5 , Beam pumping 

Leavitt . 






1 6 i „ „ . . 







[ ; Triple 

; 1 Worthington . 







Low duty 

' (Grand Junction) 

2 Worthington . 

Parkes . 






High duty 

(Thames Ditton) 
' .3 Worthington* 

Unwin . 






4 Worthington . 

Chad wick 





1-66 1 

(West Middlesex) 

5 Allis pumping 



121 203 






The figures in square brackets in the last column but one are estimated on the 
aasnmption that 9 lbs. of steam are produced per pound of coal. This would be for most 
cases about 10*7 lbs. from and at 212'' F. 

* On a lift of sixty feet only. 



Table II.c— Condensing Engines 
Stationary JBnffiim 



Tjitt of engine 

I Authority S 'r 

I i.r 




(Soath Kensington) 

Corlias (Prague) 
Corliss (Paisley) . 



(South Kensington) 

I Compound 

. I Beam pumping 
I (Copenhagen) 

Beam (Hammersmith) 

Inverted pumping . 

Unwin | 
Doerfel { 


ll)>«. per 
sq. in. ft. per 
gange miii. 
411, 60-6 412 
59-6 373 
59-2 651 
62-2 54H 

60 .">20 

61 621 



26 69 



' lbs. ! 

, 3*6;) I No jacket 
' 2-94 ■ Jacket 
— I No jacket 
No jacket 

TT««^« f ' ^4-1 66-7 343 2106 
^°^° 1. 45-6; 67 8 352 19-62 

Pumping Snffinet 
I ,1 65-8 52-5 rjf 2 23.84 

'i 81-2 61-n 264 1 19-" 
'Mair. (^«2 49-7 1 ^««:« 18-2 
' R-nJey I 168 45)0 ' ilj? 16M 

Davey&( 140 i:i0 138 1722 
Bryan "i 138 130 1374 1646 


Table II.d. — Condensing Engines 

Stationary Enffine$ 

2-30 ' No jackets 
213 Jackets 

No jackets 



! No jackets 


No jackets 



5 *? 




Type of engine 


' f Ok 

II , 





lbs. per 
n\. in. 

ft. per 




Beam condensing . 





21-51 1 

> Saturated 


»» ft • • 





19-41 1 


> steam 


»» i» • • 







1 Superheatei^. 
1 steam 


»» »» • • 










Hor. condensing (Colmar) 

Unwin . 










•» ♦ . • . 



1 99 




> Superheated 
r steam 


M . . • . 



, 94-0 


1561 , 



Taking the most favourable results which can be regarded 
as not exceptional, it appears that in test trials, with constant 
and fall load, the expenditure of steam and coal is about a& 
follows : 

Per iiidicnted lu-p. Per effectlye iL-p. 

I hour I hour 

Coal ! Steam Coal | Steam 


lbs. I Ib9. Ibe. I Ibe. 

Non-oondensiDg engine . . . 2*20 200 ' 2*44 . 220 
Condensing engine ... .1 1-60 13-5 1-76 ' 16-8 

These may be regarded as minimum values, rarely sur- 
passed by the most eflScient machinery, and only reached with 
very good machinery in the favourable conditions of a test 

It is much more diflScult to get the consumption of coal by 
engines in ordinary daily work. What is known shows that 
the consumption is greater than in engine trials. Some com- 
paratively large pumping engines, which work with a steady 
load night and day, and which worked with 2 lbs. of coal per 
effective or pump h.p. on a test trial, used 2-7 lbs. in ordinary 
working. The consumption was measured over many weeks, 
during which they were working 90 per cent, of the whole time. 
Here the consumption in ordinary work is 35 per cent, greater 
than in a test trial. 

The large pumping engines of the Hydraulic Power Company 
are rather less favourably circumstanced for economy. They 
gave an i.h.p. on trial with 2*19 lbs. of coal per hour. In 
ordinary work they are stated to use 2*93, or about 35 per cent, 
more. These engines have a fairly steady load during the day 
and a smaller load at night. 

If such a case as that of an electric lighting station is con- 
sidered, where the load fluctuates very greatly, the maximum 
load being often four times the mean load, and the minimum 
load one twentieth of the mean load, then the consumption per 
h.p. is very much greater. Mr. Crompton has given the figures 
for the Kensington station, which has excellent Willans com- 
pound non-condensing engines. Those engines will work with 



2 lbs. of coal per eflFective h.p. hour in trials at full load. 
I*esults obtained in ordinary working were as follows : 


Table III.— Coal Used in Lb& peb Houb 


Per electrical 
uuit generateil 

Per effectiye 

Per indieatal 

4 35 







In the discussion on Mr. Crompton's paper instances were 
given of coal consumption at electrical stations still larger than 
any of these. Probably up to the present the consumption has 
in no case been less than 6 lbs. per unit generated ; 3*8 lbs. per 
eflTective h.p. ; or 3-3 lbs. per indicated h.p. This large con- 
sumption will be traced later to two classes of waste, engine 
waste and boiler waste, due both of them to the ineflBciency 
caused by variation of load.^ 

In the case of small isolated motors, not generally of very 
good construction or well proportioned for their work, still more 
extravagant results have been observed. The followiug table 
gives some results obtained with small workshop engines in 
Birmingham : 

Table IV.— Coal Consumption peb Indicated H.P. Houb 
IN Small Engines at Bibmingham 



i.b.p. at full 


Actual ttvcrafre i.h.p. 
during the 

Coal consamptiou in | 
Ibfl. per i.lj.p. houi : 
durinK the | 








' The latest Board of Trade returns from electric-lighting stations confirm 
these figures. Taking the largest and best stations, the consumption of coal 
vanes in different cases from 7 lbs. to 12 lbs. per unit of electricity generated. 
It is more than this per unit sold. 


These last results are interesting both as showing how large 
fuel waste may be in unfavourable conditions, and also for this 
reason, that it is such uneconomical small engines which are 
displaced when central-station power distribution is introduced. 
It is because these small, badly-loaded engines are so extravagant 
that power can be distributed from a central station at a profit. 
As to the case of electric light stations, seeing that they are 
central stations of the type specially considered in these lectures, 
it is desirable to analyse more in detail the causes of waste. 

The Chief Secoiidftry Loss or Waste in the Action of Heat 
Motors, — The range of temperature, between the temperatures 
at which the working fluid is received by and discharged 
fix)m a heat motor, is fixed by circumstances over which the 
engineer has little control. Thermodynamics show that for 
any given temperature range there is a limit to the possible 
eflSciency of the motor. Of Q nnits of heat given to a heat 
engine, working between the temperature limits Tj and t^ 

(absolute), the part Q -* ? may be converted into work, but 


a part which cannot be less than q ^^ must be wasted. This 

_ T, 

practically unavoidable loss, arising out of the conditions under 
which heat and work are convertible, may be termed the primary 
Ices in a heat engine. For steam engines, at most three-eighths 
of the heat given to the steam can be converted into work. In 
gas engines perhaps one-half might be. 

Practically, no heat motor even approximately reaches so 
good an efficiency as this. There are secondary causes of waste, 
some of them important. Part of these secondary losses are due 
to bad construction or bad management, and can be consider- 
ably reduced by known arrangements. Part, however, are 
practically unavoidable, or at all events can only be partially 
obviated. The principal cause of the largest secondary waste 
of heat energy is essentially the same in steam engines and gas 
engines ; it is a consequence of the enclosure of the working 
fluid in conducting metallic walls. In steam engines, water on 
the cylinder walls evaporates during exhaust, cooling the walls. 
The walls have to be re-heated during admission by the con- 
densation of fresh steam. Nearly all the heat of the steam so- 
condensed is wasted, during the next period of exhaust, by re- 


•evaporation durmg the return stroke, when no useful work is 
done. The exhaust waste increases with the ratio of the area 
of admission surface of the cylinder to the weight of steam 
admitted. It increases, therefore, for light loads, because more 
surface is exposed during admission per pound of steam ased. 
It is greater when the engine works slowly than when it 
works fast, the piston effort being the same. The evil can be 
•diminished, but not entirely overcome, by steam-jackets, or by 
superheating. It is an evil specially prejudicial when engines 
are used which are too large for the work to be done. 

In gas engines it is necessary, to prevent destruction of the 
•cylinder by the high temperature of the burning gases, to enclose 
it in a water-jacket. M. Witz has shown that it is due to the 
cooling action of this water-jacketed wall that part of the gas is 
kept below the temperature of combustion. The jacket there- 
fore diminishes the efficiency of the engine, not only by directly 
abstracting heat, but by preventing the full development of the 
gas pressure early in the stroke. As in steam engines, the evil 
is greater the greater the wall surface exposed at the moment of 
•explosion. This appears to be the principal reason why initial 
•compression of the gases is necessary for good efficiency. The 
gases reduced by compression to a smaller volume are exposed 
at the moment of ignition to a smaller area of cylinder wall. 

It is useful to get a clear numerical idea of the relative 
importance of the cylinder wall action and the other actions 
•during a stroke, for this cylinder wall action is the principal 
factor in the inefficiency due to variable load or arising out of 
the use of underloaded engines, both matters of importance in 
•considering the advantages of distribution of power. Professor 
Dwelshauvers Dery has shown ' how the heat exchange, between 
the steam and the cylinder wall during the stroke, may be re- 
presented by a diagram on the same scale as the indicator 
diagram. Fig. 2 shows such a diagram draw^n for the data of 
•one of Mr. Mair Bumley's engine trials. The engine was a 
single cylinder beam engine, with jacket, working at about 4^ 
expansions, and furnishing 123 indicated h.p. The total steam 
used per stroke was 1*14 lb., or 31 cubic inches of water. The 
whole of this if condensed and spread over the cylinder wall 

» Investigation of the Heat Eaopenditure in Steam Engines ; Piroc. Inst, of 
<3ivil Engineen, vol. xcviii., 1889. 



would make a layer less than one-hundredth of an inch thick. 
The I'ange of temperature, between the initial steam temperature 
and the exhaust temperature, may be taken roughly as 200°, and 
about 30 per cent, of the steam was condensed during admission. 
The whole heat of this initially condensed steam would only be 
sufficient to heat a very thin layer of the cylinder wall from the 
exhaust to the admission temperature. 

Mr MAms Thtal L 

ifUMdml Cotvi*PhMaitLOfv jt^d fi*'*rapm^^iAtjrnr 


The dark shaded line in the figure is the indicator diagram 
of the engine. The saturation curve shows where the expansion 
line of the diagram should have been if there had been no con- 
densation. The two shaded areas represent to the same scale 
as the indicator diagram the heat given to the cylinder wall 
during admission and compression, and abstracted from it during 
expansion and exhaust. They represent a quantity of heat first 
abstracted from the working fluid and finally wasted. It will 


be seen easily that the cylinder wall action involves a larger 
quantity of heat than the whole of the heat employed in doing 
■ useful work. 

Methods of Diminishing Cylinder Condensation, — ^There are 
three ways in which the prejudicial action of the cylinder 
wall is combated. The first is the use of a steam-jacket, the 
temperature in which is higher than the mean temperature in 
the cylinder. The jacket, therefore, supplies heat to the 
cylinder wall during the stroke and lessens the amount of heat 
which must be given by the steam to the wall during admission. 
Table II.C above shows how advantageous in most cases the 
use of a steam-jacket is. But though a steam-jacket reduces, 
it does not prevent initial condensation. Further, the greater 
the speed of the engine the less time there is for heat from the 
jacket to penetrate the cylinder wall, and the less efiective the 
action of the jacket becomes. The second method of diminish- 
ing the cylinder wall action is to carry out the expansion of the 
steam in stages; that is, to use compound or triple engines. 
Then the temperature range in each cylinder and the admission 
surface per pound of steam in each cylinder are less, so that to 
some extent condensation is directly diminished. But also the 
steam re-evaporated during exhaust in one cylinder forms part 
of the admission steam to the next cylinder. Tables II. a and b 
show clearly the advantages of stage expansion. There is yet 
one other method of reducing the action of the cylinder wall, 
and that is to use superheated steam. Table II. D. The cylinder 
wall is then to a considerable extent reheated by the superheat 
in the steam without condensation. 

• Him discovered, forty years ago, that when steam is heated 
above its saturation temperature, or superheated, before ad- 
mission to the cylinder, the initial condensation is diminished 
more efiTectively than by jacketing. In the years 1854 to 1865 
superheating was somewhat extensively used, especially in the 
Navy, and always with a marked gain of eflSciency. Various 
practical difficulties in the use of superheaters, especially the 
danger when wrought-iron superheaters were overheated, led to 
the disuse of the process. 

Lately superheating has been reintroduced in Alsace, where 
its advantages were first discovered, and superheaters have been 
applied to a very large number of boilers. Many experiments 



have been made by Alsatian engineers with saturated and super- 
heated steam in the same engines. In all cases they have 
found an economy ranging from 10 to 25 per cent, when 
superheated steam was used. It has been commonly alleged 
that the high temperature of superheated steam causes scoring 
or erosion of the cylinder and valve faces. Such injury did 
occur in the early use of superheated steam, for at that time no 
lubricant was obtainable capable of standing a high temperature. 
But this danger has probably been very greatly exaggerated. 
The cooling action of the cylinder is so great that superheated 
steam does not retain its high temperature for a sensible time 
after admission. With ordinary care and the use of a good 
lubricant, it does not appear that the engines using superheated 
steam suffer any injury. 

liately, the author had an opportunity of testing a large 
mill engine in Alsace using superheated steam. The engine 
was a horizontal receiver compound engine, each cylinder having 
four slide valves and Corliss g^r. The cylinders were steam- 
jacketed, and were used with steam in the jackets in all the 
trials. The normal steam pressure was 85 lbs. to 100 lbs. per 
square inch. The following table gives the principal results : 

Tbials op Mill Engine at Logelbach with Saturated and 
supebheatbd steam 

TiiallL , Trial L 

With satu- With guper- 

' rated steam heated steam 

Total indicated h.p. 

Boiler pressure, lbs. per sq. inch . 

Amount of superheating 

Pounds of steam per lb. of coal 

Pounds of steam per i.h.p. hour 

Pounds of coal per i h.p. hour 

Per cent, economy of steam due to super- 

Per cent, economy of coal due to super< 

Trial VI. 
With super- 
heated steam 




9905 1 


118°-3 F. 



' 19-76 





20 9 






The coU was of poor quality. 

The superheaters used in this case were constructed of cast- 
iron pipes of special form, under patents taken by M. Schwoerer 
of Colmar. The superheater is usually placed in a chamber 
forming part of the boiler flues ; sometimes a detached super- 
heater with separate fire is used. 





Pigs. 3, 3a, show the general ari*angement of a superheater on 
M. Schwoerer's system. The boiler in this case is an elephant 
boiler. The cast-iron coil which forms the superheater is placed 
in a chamber below the 
boiler, through which the 
furnace gases are taken 
before they are too much 
cooled by contact with the 
boiler surfaces. The super- 
heater requires no attention 
while the boiler is working. 
Its construction is such that 
there is no reasonable pro- 
bability of injury or danger 
from overheating. 

Ijoad Curve ai\d Load 
Factor, — ^A curve, the ab- 
scissae of which represent 
time, and the ordinates 
the rate of expenditure of 
•energy, is called a load 
curve, and such curves are 
very commonly drawn for a 
period of twenty-four hours' 
working,becau8e,apart from 
seasonal fluctuations, a day 
is a natural period in the 
operation of a power plant. 
The ordinates may be horse- 
power, or volt-amp6res, or 
units of any other quantity 
proportional to the rate of 
expenditure of energy. The 
area of the curve represents 
the total amount of energy 
for the period considered. 
A load curve may be drawn 
for a single engine or 

machine, or for a plant of many engines or machines. The load 
line for a central station is that to which attention is to be 


Fig. 3 a. 



directed. The ordinate of each a station load curve represents 
the sum of the energy expended at the moment per unit of time 
by all the engines or machines then in operation. 

Load curves for particular cases have, no doubt, been 
frequently drawn, and the influence of fluctuation in the rate of 
working on economy has been noted. But it is due to Mr. 
Crompton that the use of the load curve, in examining the results- 
of station working and in discriminating the causes of differences- 
in the results obtained in diffirent stations, was first clearly 

LQAO CURVES lonoon hydraulic pow£R stat/on 

my cyMv£ 


Fig. 4. 

LOMO rmero* • '420 

COAO fACT0ff4&0 

It can be directly inferred from the examples given in Mr. 
Crompton's paper that the cost of working per unit of mechanical 
or electrical energy distributed, in different electric lighting 
stations, depends very intimately on the form of the load curve. 
Mr. Crompton introduced the term * load factor ' to express the 
coeflScient of fluctuation of the rate of working. There may be 
various load factors, according to the precise fluctuation con- 
sidered. But for the object at present in view, the considera- 
tion of the influence of variation of load on the eflSciency of 
* Electrical Energy ; Proc. Inst. Civil Engineers, vol. cvi. 



«team plant, the load factor may be taken to be the ratio of 
the area of a day's load curve to the area of a rectangle 
■enclosing it. It is equally the ratio of the average load during 
the day to the maximum load at any time during the day. 
The plant must be large enough for the maximum load. 
The income depends on the amount of energy delivered. The 
efficiency of the engines depends on the load factor. The cost 
of a day's working depends partly on the average output, partly 
on the load factor. 


$ ^UN£ IB90 rtNt .- ns*mAM,mm^». 

Jt 0£C, /8S0 FOG . '6Z 

9 *fAN. 199/ nN£ ' .#/ 

Fig. 5. 

Pig, 4 gives two load curves for a day's working of one of 
the stations of the London Hydraulic Supply Company. These 
indicate the kind of fluctuation of demand which occurs in a 
central station supplying power for a large number of inter- 
mittently working machines, chiefly lifts and hoisting machines. 
Such machines are in frequent use in the day, and are little 
used at night. The demand for power-water pumped by the 
engines at the station is large and pretty constant from 9 A.M. 
to 5 P.M. During the remaining hours the demand is small. 
The load factor for the day, understood as defined above, is 0*42 



in 1887 and 0*46 in 1891, when the system had been consider- 
ably extended. This shows that as the number of consumers 
supplied is greater, the demand is a little more uniform. 

Fig. 5 shows load curves for the London Gas and Coke 
Company. A gas generating station is essentially a centi'al 
station supplying and distributing a means of producing energy 
either for lighting, heating, or power purposes. The ordinate^ 
in this case represent cubic feet of gas supplied per hour. If, 
say, 26 cubic feet of gas per hour is assumed to be capable of 






/ .••' 

























> ■ ■ 




LOAD CURVES lanvmroN coort elcbtric station. 

Fig. 6. 

furnishing a horse power, it is easily seen that the ordinates of 
the curves to a suitable scale represent equally horses' power of 
energy supplied. In the case of the Gas and Coke Company, 
the largest demand is for lighting, and this is greatest in the 
evening. But there is also a considerable demand during the 
day for gas for heating and for power. The daily diagram 
factor was 041 for a day in January, falling to 013 for a day 
in June. On a foggy day in December it rose to 0*52. 

Pig. 6 gives load curves for the Kensington Electric Light- 
ing Station. As practically the whole of the electricity 


generated is used for lighting purposes, the period of large 
demand is short, and the fluctuation of demand greater than in 
either of the previous cases. The daily load factor is 0*24 for one 
of the curves, and 0*31 for the other. But for the partial use of 
storage batteries, the load factor would have been smaller stiU. 

Indicated and Effective Horse Pover, — In questions of 
power distribution it is clear that it is the effective horse power 
delivered at the crank shaft, and not the indicated horse power 
developed in the cylinder, which has to be considered. It is due 
to the difficulty of determining in most cases the mechanical 
efficiency of an engine that engineers have been content to reckon 
on the indicated horse power. It is true that the engine fiiction 
is not a very large fraction of the power developed, in full load 
trials, nor does this fraction vary very greatly at full load for 
different engines. But it is erroneous to assume tacitly that 
the engine friction is in all cases a quantity of relatively little 
importance, or that it is immaterial whether the steam con- 
sumption is reckoned on the indicated or the effective horse 
power. As the engine friction is nearly the same at all loads, 
then, though it is only a small fraction of the indicated power 
at full load, it is a large fraction at light loads. The electrical 
engineer who uses engine power and has exact means of 
measuring the quantity of power delivered to dynamos, veiy 
naturally and rightly pays more attention to effective than to 
indicated power. 

Influence of Mechanical Efficiencif on the Economy of Work- 
ing tcith a Varying Load. — ^The mechanical efficiency of steam- 
engines, or the ratio of the effective to the indicated power 
at full load, is 0-8 to 0*85 for small engines, and may reach 
at least 0-9 for large engines. It is a little greater for non- 
condensing than for condensing engines, and for simple than 
for compound. A triple expansion ergine constructed by 
Messrs. McLaren, tested on a brake, gave 122 i.h.p., and 107 
on the brake, an efficiency of 0-88 ; a very good result for so 
complicated a machine as a triple engine necessarily is. The 
loss of power due to engine friction is not very great, or even 
for different types very variable, so long as the engines are 
worked at full load. It is quite otherwise, however, at light 
loads, and the extent to which this affects the economy of work- 
ing has been overlooked. 



Many experiments show that the engine friction is nearly 
the same at all loads. 

Let T^ be the effective h.p. 
T^ the indicated h.p. 
F the engine friction in h.p. 

T, = Tf — F 
and the efficiency is 

i? = t,/t,=t,/(t, + f) 
The following Table gives the results of some experiments 
on a, small non-condensing engine : 

Mechanical Efficiency of a Small 
Non-condensing Engine 

ludicated b.p. 

Brake h.p. I of full indicated 







' 8-86 



























The experiments were made at different times, so that there 
is a little irregularity in the results. The results agree approxi- 
mately with the equation T^=0-95 T, — I'-i. 

It appears, therefore, that the assumption that engine friction 
is independent of the load is sufficiently approximate. Suppose 
an engine works at 100 indicated h.p. and 85 effective h.p. at 
i'uU load. Its efficiency at full load is then 0*85. At other 
loads it will be as follows : 

Table v.— Mechanical Efficiency of Engines 
WITH Vakying Load 

Indicated h.p. 

EffectlTe li.p. I Efficiency 


1 85 



' 60 












The decrease of mechanical efficiency for light loads is 
remarkable, and has a serious influence on the economy of 
working with varying load. 

Careful experiments on mechanical efficiency with varying 
loads are not very numerous. It is useful, therefore, to give the 
results of some experiments on a Corliss engine of about 180 i.h.p. 
At full load. This engine was tried with a brake at Creusot, 
both condensing and non-condensing. It was found that the 
results agreed approximately with the following equations : 

Condensing t, = 0^902 t, — 16 

Non-condensing T^— 0945 7^ — 12 

equations which give results not differing greatly from those 
obtained by assuming the friction constant. The following are 
the calculated values of the efficiency : — 

Table VI.— Mechanical Efficiency of Corliss Engine 
WITH Varying Load 

Ratio of actual 

Meclmnical efficiency 

effective power 

to power 

at f uU load 


















Influence of the Loss due to Back Pressure on the Economy 
■of Steam-engine icorhinci. — Besides engine friction, there is 
another waste of energy in the steam-engine which has to 
an even greater extent been overlooked. The effective power 
is less than the indicated power by the engine friction ; but 
the indicated power itself is less than the work done by the 
steam, by the amount of work done against back pressure. 

In condensing engines the back pressure is comparatively 
small, but in non-condensing engines the back pressure exceeds 
15 lbs. per square inch. Its influence on economy, even at 
full load, is considerable, and at light loads it may become 
excessively great. 

In engines working, as most engines do, at constant speed, 
the work against back pressure is nearly independent of the 



load. In interpreting^ an indicator diagram (fig. 7) the total 
work done by the steam on the piston, called by some 

Continental writers the abso- 
lute indicated ivorJc, is the area 
a b cd. The work afterwards 
wasted in overcoming back 
pressure is o afc d. The differ- 
ence is the effective work ab cf. 
The quantity of steam used de* 
pends on the absolute indicated 
work ; the useful energy ob- 
tained on the effective indicated 
work. If the back pressure 
work is constant it becomes a 
the absolute work as the load 

Fio. 7. 

larger and larger fraction of 
on the engine is diminished. 

Suppose in a non-condensing engine the work against back 
pressure is 20 per cent, of the absolute indicated work of the 
steam at full load. Then for other loads the work is distributed 

Table VII.— Waste op Work due to Back Pbessure 


Indicated work ' 

of steam 


against back 




Net or 
effective indi- 
cated work 





From the figures in the last column, the friction has to be- 
deducted to find the useful effective work. 

The following Table is calculated from the indicator diagrams- 
of a compound engine working at nearly constant speed with a 
var}'ing load. The waste work against atmospheric pressure is- 
calculated, exclusively of the waste of work doe to excess back 
pressure, due to resistance of passages, &c. It represents^ 
therefore, work wasted in a non-condensing engine which is- 
almost entirely saved in a condensing engine : 



Table VIIL— Wobk Lost in Pumping against the 

I I \ ; 

I j Effect ire indicated work iu li.p. Work wasted 

No. of Total effective I against 

I trial I indicated work atmospheric 

' H.P. cylinder , L.P. cylinder I pressure 






The work wasted is equal to f of the effective work at full 
load and to If of the effective work at the lightest load. 

At full load a well-designed non-condensing engine doe& 
not use much more steam per effective i.h.p. than a condensing 
engine. The gieater back pressure in the former is partly- 
balanced by the greater cylinder condensation in the latter, due 
to the greater temperature range. But with light loads it is 
very different. Hence condensing engines should be used, if 
possible, whenever the load is a very varying one. 

It is stated that at the electric station at Gothenburg the 
fuel consumption was reduced by sixty per cent, when con- 
densers were added to the engines. This economy was obtained 
in spite of the fact that at full load the engines worked nearly 
as economically when non-condensing as when condensing. 

If the power of an engine is varied by vaiying its speed, 
instead of by varying the work per stroke, the speed being 
constant, then the conditions are different. If the power i» 
varied by varying the speed, then the work against back 
pressure bears the same ratio to the effective work at all loads. 

Influence of the Type of Enghie, the Speed, mid the Mode of 
Regulation on the Thermal Efficiericy, — It has already been 
pointed out that there are very large thermal losses in 
heat engines which are not shown on the indicator diagram, 
and which have a very important effect on economy of 
working. Those thermal losses are greater also at light loads 
than at full load, and they vary very much with the type of 
engine and some other conditions of working. Unfortunately, 
there is not a great deal of information available as to the 


steam consumption of different engines, except in the case of 
full load trials. There are the experiments of the late 
Mr. P. W. Willans, to which reference will be made presently. 
But, except these, there are very few which afford much guidance 
as to the l-elative economy of engines working with varying 

It is possible to get a good idea of the influence of conditions 
of working on the thermal eflSciency in this way. Professor 
Cotterill * has found a means of calculating the cylinder con- 
densation in an unjacketed simple engine. The rest of the 
steam used can be ascertained in other ways. By examining 
the steam consumption in a variety of conditions for such an 
engine, a good deal of insight may be gained applicable to all 
cases. With the aid of Professor CotterlUs formula the steam 
consumption has been calculated for a number of cases and 
the results plotted in curves. Some check on the general 
bearing of these results can then be obtained by plotting in a 
similar way such experimental results as are available. 

An engine has been assumed working at full load in given 
conditions. Then the effect on the steam consumption of 
varying the speed, the initial steam pressure, and the ratio of 
expansion has been calculated. The results have been plotted 
in curves which give the steam consumption in pounds per 
i.h.p. hour at any fraction of full load. 

Let N = revolutions per minute 

d = diameter of cylinder in feet 
r = ratio of expansion 

is the condensation per pound of steam admitted to the cylinder. 
Let j>i be the initial (absolute) steam pressure and i\ the back 
pressure in pounds per square inch ; i\ the volume in cubic feet 
of a pound of steam at j)i. Then the indicated work done on 
the piston per pound of steam not condensed is, — 

liipit'jh+log. r-''^Aj 

In consefjnence of condensation the work per poand of steam 
' The Steam Evfjiine, 1890. chap. xi. 


actually used is reduced in the ratio 1 to 1 + ?' . Hence^ 

in an engine in which condensation occurs, the effective work 
per pound of steam is 


144 j9i r, i^K 

1 + ^^??-J 

But a h.p. hour is 1,980,000 ft. lbs. of work. Consequently 
the number of pounds of steam required per i.h.p. hour will be 


w = 

144;7,v, jl + log. r^Lhl 

For the following calculations the engine has been assumed to- 
have a cylinder four feet in diameter, and the constant c has 
been taken equal to six. In considering the effect of speed, 
a slow engine working at 12^ revolutions per minute, an 
ordinary engine working at 25 revolutions per minute, and & 
fast engine running at 50 revolutions per minute, have been 
assumed. The back pressure is taken at 3 lbs. per square inch 
for condensing, and at 16 lbs. per square inch for non-condensiug 

Methoih of Eegulation when the Load varies, — There are 
three ways in which the conditions of working may be varied 
when the power demand varies. The speed may be varied, as i& 
often done in the case of pumping engines. The initial steam 
pressure may be varied, which alters the weight of steam used 
per stroke by altering the density of the steam. This may be 
done by varying the boiler pressure or by throttling the steam. 
Lastly, the expansion may be varied. A case of an engine has 
been taken and the effect on the steam consumption of these 
different ways of varying the power has been calculated. Figs. 
8 and 9 show by curves the results both for condensing and non- 
condensing engines. The results are theoretical, but they have 
been compared with data from various engines, and they agree 
with them quite closely enough for the purpose of comparison. 
Strictly, however, these results are applicable only to unjacketed 



Case I. — Power Varied hy Varying the Speed. — p, s=110 
lbs. per sq. in.; Vi=s3'99 c. ft.; normal ratio of expansion 4. 
The steam consumption for these conditions per i.h.p. hour is 
given in fig. 9. As the speed changes from 50 revolutions at 
full load to 6^ revolutions at 12^ per cent, of full load, the steam 
consumption rises from 18 lbs. to 25 lbs. per i.h.p. hour. 

Case II. — The hiitial Pressure Varied. — Normal ratio of ex- 
pansion 4 ; p, at full load 110. The results for a fast, ordinary, 
and slow engine are shown in fig. 9. 

Case III. — Ratio ofExpaiision Varied. — p, = 110. The steam 
consumption is calculated for a fast, an ordinary, and a slow 
engine. The ratio of expansion at full load is taken at 2 only. 
The results are given in fig. 9. It is due partly to the small 
•expansion assumed for full load that the steam consumption per 
i.h.p. diminishes as the load diminishes. It would increase with 
small loads if difierent assumptions were made. 


The same cases with the same data are calculated and the 
results given in fig. 8. Only, for the case of varying expansion, 
the ratio of expansion at full load is taken at 4. 

The results are plotted in the curves shown in figs. 8 and 9. 
These curves cannot be taken as absolute guides, partly because 
a constant value for c has been taken, partly because the 
formula is only trustworthy within limits. Still, they are very 
instructive as indicating the way in which variation of load 
causes a variation in the steam consumption. 

Curves for Actual E'ngines Similar to the Theoretical Curves. — 
In order to test how far the curves just given agree with tests of 
actual engines, curves drawn in the same way for some engine 
trials in which the engine was tested at different loads are given 
in figs. 10, 11, 12. It will be seen that although these curves 
embrace a remarkable variety of engines, the size, the speed, the 
type of engine, and the initial pressure all varying considerably, 
yet the curves for corresponding cases are very similar to the 
theoretical curves previously given. 













^ 6 


""^ """^S S 8" 5 % '^ ^ 

i//?0// d3d dHi did nvji% JO SONnOd 

^ 5 t 9 $ 5~ \ ^ %- 

tfnOH dJd dH't i^Jd'iVVS2S JO SOAfnOd 



Fig. 10 shows curves drawn from data in Mr. Willanss 
paper on non-condensing engines, simple, compound, and triple. 
Also one carve for a triple condensing engine. The full carves 
show the effect of regulating by varying the boiler pressure or 
by throttling. The dotted curves show the effect of varying 





C0M»€M*/t»9 rmt^tt 

P£iiG£UTdG£S Of fULL £fr£CTIVC HjR. . 

Fig. 10. 

expansion, and varying both expansion and pressure. In this 
figure the abscissae are percentages of the full load effective or 
brake h.p. Fig. 11 gives the same results, the abscissae being 
percentages of the full load indicated h.p. The comparison of 
the two diagrams is instructive, because it shows how much 



more rapidly the steam consumption increases at light loads 
when the real useful or effective work is considered than when 
the indicated work is considered. It shows that to pay attention 
to the indicated work only is misleading. 



KIND or €NCM£ 

VARIA TlON 0,r / HP «v 


trnt-cofottiiif rmt»it 
t-iofn,— rm.m^t 



f.. ........ 




























^"^"' ■ 






• 4 

> or r 



? HOR 





<f «. 


Fig. 11. 

Fig. 12 shows a number of miscellaneous results selected 
from the few cases in which large engines have been carefully 
tested with varying loads. The results are more irregular than 
Mr. Willans's results, as might be expected, but they are 
instructive nevertheless. 




Mr, WiUans's Law. — In the discussion on Mr. Crompton's 
paper* the late Mr. P. W. Willans first stated a remarkable 
simple approximate law for the total steam consumption of an 
engine, working at constant speed with a constant ratio of ex- 




mfifA Tior^oriHP BY 




VAMTtM^ r 

• • 

V»AY*M» r 

• 2 


» • 


I £1 tA* 

VAAyiM€ p, AIM* N. 



















• ' 






r — 






1 ;! 

. J 

k • 

t » 

• « 

• « 

• m 

1 ■■ - • 

• «i 



Fig. 12. 

pansion. He found that the total weight w in lbs. of steam used 
per hour for any engine in the conditions stated was given by a 
linear equation of the form 

ti; = a + 6h.p. 

where h.p. is ihe horse-power at which the engine works. Thus, 

* Ptoc, iTUt. Civil Engineerij vol. cvi., p. 62. 



for an engine of 100 indicated h.p. at full load, he obtained the 
following equations, in which i.h.p. is put for indicated h.p., and 
•e.h.p. for electrical h.p., which is taken at 80 per cent, of the 
indicated power. 

Nmi-condensing Triple, — (About 6-7 expansions.) 

t(; = 450 + 13-75 i.h.p. 
= 725 + 13-75 e.h.p. 

Non-condensing Compound, — (About 4*45 expansions.) 
10 = 525 + 16-25 i.h.p. 
= 850 + 16-25 e.h.p. 

Condensing Triple. 

w = 112 + 13-75 i.h.p. 
= 377 + 13-75 e.h.p. 

If, instead of calculating the total steam per hour w, we calculate 
the steam per h.p. hour, we get results which, plotted like the 
previous diagrams, give rectangular hyperbolas, curves which 
a^ee closely with the theoretical curves for the case of varying 
pressure previously given. To show the great variation of steam 
consumption these values are given in the following tables for 
fall load, half load, quarter load, and one-eighth load : — 

Table IX.— Steam Consumption, Lbs. pbb I.H.P. Houb 

Indicated h.p. 









Steam Consumption, Lbs. pee Electbical H.P. Houb 























Although these results are obtained from Mr. Willans*9 
formula, they agree with remarkable exactness with his experi* 
mental results, and may be taken to be experimental results 
obtained with extremely good engines. Nothing could more 
strikingly exhibit (1) the decrease of efficiency at light loads, and 
(2) the very gi'eat superiority of the condensing engine at light 

Lun-ea^e of Steam GoTisumption Worldiig with a Variable 
Load, — Captain Sankey has applied Mr. Willans's formula 
to find the steam consumption of one or more engines 
working against a variable load, as in an electric lighting 
station. He takes a normal midwinter load curve and examines 
how the necessary current could be supplied during the twenty- 
four hours : (1) with an engine capable of exerting the 
maximum power required, (2) with smaller engines. He also 
considers the steam consumption when one additional engine ia 
kept running at half speed as a stand-by in case of accident. 
The results, rearranged and a little modified, are given in the 
following table. It is assumed for convenience that the 
maximum load is 500 electrical h.p., and that the engines are 

Table X.— Stkam Consumption in Engines Working with a 
Yabiable Load 

Steam oonsump- 

Per cent, increaae 


tion in lbs. per 

of steam con- 


load factor 

ayerage electrical 

sumption due to 

h.p. hoor 

variable load t 


I. 600 e.h.p. enKine . 




la. 500 e.h.p. engine, and 

Himilar engine running at 


half speed 




II. 200 e.h.p. engines 




Ila. 200 e.h.p. engines, with 

one similar engine run- 

ning at half speed . 




III. 100 e.h.p. engines 




Ilia. 100 e.h.p. engines, and 

one similar engine run- 

ning at half sp^d . 




Injinence of Irregular W&rTdiig of the Boilers mi tlie 
Expenditure of Fuel. — With a varying load the steam con- 
sumption, and consequently the fuel consumption also, is 
increased: (1) in consequence of the decreased mechanical 



eflSciency of the engines with light loads, (2) by the greater 
proportion the work expended in overcoming back pressure 
bears to the total work of the steam, (3) by the diminished 
thermal efficiency of the engine. But all these causes taken 
together do not explain fully the great fuel consumption in such 
i^ases as electric lighting stations. There is another very 
obvious cause of uneconomical working which cannot at present 
be estimated quantitatively for want of sufficient experimental 
investigation. With a very varying load boilers must be put 
in steam and banked up alternately, and the waste in getting 
up steam and allowing the boilers and brickwork to cool down 
again is no doubt considerable. This waste is at present 
unavoidable, except so far as means can be adopted to improve 
the load line. 

Some tests made by Professor Kennedy, at the Millbank 
Street Station of the Westminster Electric Supply Corporation, 
indicate pretty clearly a boiler waste additional to the engine 
waste. Dividing the day into three portions, he determined 
the fuel consumption, the feed water evaporated, and the 
indicated and electric h.p. developed during each period. It 
will be seen that during the periods of light loading the fuel 
consumption per h.p. hour is very large. 

Table XL— ^Coal Coksumption in Boilebs with a Vabiable Load 

11 a.m. to 
6 pan. 

6 p.m. to 

to 11 a.m. 

Mean for 34 


7 hours 

6 hours 

11 hours 

Total i.h.p. hours 





Total ch.p. hours 




Ck)al per i.h.p. hour lbs. 





Coal per e.h.p. hour lbs. 





Lbs. of water evaporated 

per lb. of coal 





Lbs. water per i.h.p. hour . 





Lb6. water per e.h.p. hour . 





Perhaps the fairest way of considering the waste due to 
variable load will be to compare the mean consumption in the 
twenty-four hours with the consumption between 6 p.m. and 
midnight, when the load was heaviest. It will be seen that the 
mean steam consumption per electrical h.p. hour was 24 per cent, 
greater than during the period of heavy load. Bat the mean 


consomption of coal per hour was 46 per cent, greater 
than during the period of heavy load. The difference of 
22 per cent, mnst be attributed to waste at the boilers, dne to 
irregular working. During the whole twenty-four hours the 
mean evaporation in lbs. of water per lb. of coal was only 
85 per cent, of the evaporation during the period of maximum 

Variation of Efficiency and Fuel Consumption in Internal 
Furnace or Explosion Engines. — Gas and liquid fuel engines 
receive their charge at atmospheric pressure, as well as ex« 
hausting into the atmosphere. Hence in a complete cycle 
the resultant back pressure loss is comparatively small. The 
engine friction, however, is rather larger than in steam engines, 
and appears to be independent of the load. Hence the 
mechanical efficiency decreases at light loads. Also at light- 
loads the combustion is in some cases less perfect, or proceeds 
more slowly, and this is a cause of loss. It is well understood 
that gas and petroleum engines should be worked as far as 
possible at full load. At the Dessau Electric Station, which is 
worked with gas engines^ large secondary batteries are used ta 
store the surplus energy when not required for supply, and to 
obviate the necessity of working the engines at light load. On 
the other hand, engines of this type have the very great 
advantage that they can be started in a few minutes when 
required, and stopped whenever they are not wanted. There is, 
in a station worked with such engines, no loss like that due to 
irregular working of the boilers. 

From some experiments on gas and petroleum engines, the 
author obtained the following approximate equations for the 
amount of fuel used. 

Let w = fuel used per brake h.p. hour. 

p = fraction of full load at which the engine is working. 

For gas engines using lighting gas 

M? = , + 15-25 cubic feet of gas per hour. 

For petroleum engines 

w? = - + 0-6 lbs. of oil per hour. 



Tablb XII.— Fuel Consumption in Internal 
FcTBNACB Engines 

Brake load in per 

Gas engine. 

oil engine. 

cent, of load at 

c. ft. of gas per 

lbs. of oil per 
brake h.p. hour 

full power 

brake h.p. hour 














66 45 


It will be seen that the cost in fuel per h.p. hour increases 
greatly at light loads. 

The Steam Turbine, — It is very many years since Girard 
predicted that ultimately steam turbines would supersede 
reciprocating engines. Although many engineers have ex- 
perimented on steam turbines, Girard's prediction is still 
unfulfilled. But important progress has been made in the use 
of steam turbines. The electric stations at Cambridge and 
Newcastle are worked with these motors, and the steam turbine 
may have an important part to play in central station working. 

Credit is due to Mr. C. A, Parsons for first completely 
facing the mechanical difficulties involved in designing a steam 
turbine. Of these the most obvious is the excessively great 
number of rotations at which any steam turbine must run. The 
circumferential speed of any turbine must depend on the velocity 
due to the pressure head of the fluid used. Taking steaiA of 
140 lbs. absolute pressure, the weight of a cubic foot is 0*3148 lbs. ; 
the height due to the pressure is (140 x 144)/0-3148 = 64,000 
feet; the velocity due to this height is roughly 8-02^^64,000 = 
2,030 feet per second. It is at some large fraction of this 
enormous velocity that the circumference of a steam turbine 
must run if any good efficiency is to be obtained. 

Mr. Parsons partially met the difficulty by using multiple 
turbines, so as to divide the pressure difference to be dealt with 
into stages. In his earliest form of turbine, a series of axial 
flow turbines were placed on a common shaft, thirty or forty in 
number, with fixed guide vanes between each pair of turbine 
rings. Several hundred of these axial flow multiple turbines 
were constructed for electric light work, the aggregate horsa 



power amounting to 4,000. The speeds of rotation were 10,000 
revolutions per minute and upwards. Mr. Parsons has now 
modified the construction of his turbine and made it a radial 
outward flow turbine. The turbines in this type are arranged in 
concentric rings on a rotating disc, with fixed guide vanes between 
each pair of rings. The steam works its way 
radially through a series of guide vanes and 
wheels outwards, and then passes to the centre 
again to flow through another series. Professor 
Jit —SBta Ewing, F.R.S., has tested one of these turbines of 
'"H — JK 135 electrical h.p. The consumption of steam 
was 27*6 lbs. per e.h.p. This corresponds to 
a reciprocating engine working with 20 lbs. of 
steam per i.h.p. Since then still better results have been 
obtained, and the remarkable feature is shown that the increase 
of steam consumption per h.p. with low loads is much less than 
with reciprocating engines. 

Fig. 13. 

Fig. 14. 

Another steam turbine, less generally known at present, is 
one invented by M. Gustav de Laval. Unlike that already 
described, this is a simple impulse turbine. The steam is dis- 
charged through two, four, or more nozzles, acquiring in an 
expanding mouthpiece the full velocity due to the pressure 
head, and issuing at atmospheric pressure with its energy con- 
verted into kinetic energy. It is received on the vanes of a 
single simple impulse turbine passing through it axially. 
Fig. 13 shows one of the nozzles discharging into the wheel. 
Fig. 14 shows the turbine wheel (5), on its shaft (4), and at 
(3) the pinions of a pair of spiral gears, by which the speed is 
directly reduced in the ratio 10 to 1. The complete ex- 
pansion of the steam to atmospheric pressure in the nozzle (or 
to a lower pressure if an ejector-condenser is used) is de- 
termined by the proportions of the nozzle. The velocity of 
discharge for steam of 75 lbs. pressure expanding to atmospheric 


pressure is given as 2,625 feet per second; expanding to a 
vacnnm pressure of 1^ lbs. per square inch, at 4,600 feet per 
second. The speed of the turbine is limited by the* resistance 
to centrifugal tension. In a 5 h.p. turbine the peripheral speed 
is stated to be 574 feet per second, and the number of revolutions 
30,000 per minute. One great diflSculty with turbines running 
at so high a speed is to ensure dynamical balance. In the Laval 
turbine the shaft is very small in diameter, and has a consider- 
able unsupported length. If the balance is not perfect the 
shaft bends till the axis of rotation becomes an axis of inertia. 
There are then no vibrations. The spiral gearing takes up the 
•end thrust and also runs perfectly quietly. The speed governor 
is very small, simple, and effective. 

Experiments have been made on a 50 h.p. Laval turbine by 
Professor Cederblom. The steam pressure by gauge varied 
from 108 to 122 lbs. per square inch. A Korting ejector- 
•condenser was used, and the pressure in the exhaust pipe of the 
turbine was 1*7 lbs. per square inch. The turbine developed 
-63-7 brake h.p. with a steam consumption of 19*73 lbs. per 
brake h.p. hour, and a consumption of 2*67 lbs. of coal per 
brake h.p. hour. These are extremely remarkable results. 

The important feature of such a turbine, besides its economy 
-of working, is its extreme simplicity and the absence of wearing ' 
parts. It can be regulated for light loads by stopping the flow 
through one or more of the nozzles. It ought, therefore, to 
have an efficiency with light loads little less than that with full 



The probable cost of steam power in any given case can only 
be determined by careful estimates in which local conditions are- 
taken into account. The cost of coal, facilities for obtaining^ 
water, the cost of labour, even the type of engine and character 
of the buildings required are more or less different in different 
cases. Further, the way in which the power is applied, the^ 
number of hours the engine is used per day, and the regularity 
of the load during working hours affect very much the cost. 
Certain typical cases may, however, be taken, and an average 
estimate made of the cost in such cases. The case will be taken 
first of engines used in industry and working a regular number 
of hours daily with a nearly regular load. This will afford some 
"indication as to how far motive power, supplied from centra) 
stations by some method of transmission, can be used economi- 
cally, in place of power generated locally by st^am engines. 
Then the special case of the cost of power, generated by steam 
in central stations for distribution will be considered. 

Cost of EngineSy Boilers, and Buildings, — ^With engines of 
100 h.p. or more the cost can be pretty definitely stated, and 
the total cost of engines and boilers per h.p. does not vary very 
greatly with the type of engine adopted. For if a cheaper and 
simpler type of engine is selected, then, its eflBciency being less, 
the boilers have to be larger. But with small engines the cost 
per h.p. increases very considerably, both because small engines 
are less efficient and because they are more expensive to con- 

It will be assumed for the following estimates that the total 
cost erected of engines and boilers, with pipes and auxiliary ap- 
paratus and such buildings as are necessary, may be taken to be 
as follows : — 

Tablb XIII.— Cost op Steam Plant 


Indicated horse-power 





Sffective h.p 

Cost per i.h.p. in £ 

Cost per effective h.p. in £ . 





In determining the annual cost interest will be taken at 5 per 
cent., and maintenance (repairs) and depreciation at 7^ percent. 

Cost of Coal and Petty Stores. — In the following estimates 
coal will be taken at 20s. per ton. The amount of 
coal required must be calculated so as to allow for lighting up 
boiler furnaces, for waste due to cooling of boilers and brickwork 
when steam is let down, and for working auxiliary apparatus 
such as feed pumps. 

Table XIV.— Working Cost of Steam Plant 

Indicated horse-power 


10 60 


Effective h.p 

Coal per i.h.p. hour, lbs 

Coal per effective h.p. hour, lbs. . 







The cost of petty stores will be taken at 0'2bl. per 
effective h.p. per annum in the case of moderately large engines 
working ten hours a day. In other cases a proportionate 
estimate will be made. 

Cost of Labour. — For driving, stoking, and cleaning, an 
allowance of 1*2/. per annum per effective h.p. for 3,000 hours, or 
0-6Z. per annum for 1 ,000 hours, will be made. In the case of 
engines of 10 h.p. or less, however, the labour reckoned on the 
h.p. costs considerably more. 

Table XV.— Cost op an Effbctive H.P. per Ybab op 1,000 Woekinq 
HouBS. Thb Engine Wobkino Reoulablt with neably Full Load 

Indicated horse-power of engine 

Interest at 6 per cent, on engines, boilers and 

Maintenance and depreciation at 7^ per 
cent. .'....:.. 

Coal at 20». per ton 

Petty stores ....... 


Total cost of 1 effective h.p. per year of 

1,000 hours in £ 

Cost in pence per effective h.p. hour 

































Table XVI.— Cost of an Effective H.P. pbb Teab op 3,000 Work- 
ing HouEs. The Engine Wobking begulably with neably Fcll Load 

IiitlioAted horae-power 

Interest at 5 per cent, on engines, boilers and 
buildings | 400 

Maintenance and depreciation at 7^ per 

Coal at 20s, per ton 

Petty stores 


Total cost of an effective h.p. per year of 

3,000 working boars in £ . 
Cost of an effective h.p. hour in pence . 





























! 38-64 








The results given in these tables are plotted in figs. 15 and 
16. The extremely rapid increase in the cost of working for 
small powers is very striking. 

y£A/9 or tooo hours 

MO^se mtrnttm 

Figs. 15 and 16. 

Annual Cost of an Kffedire H.P. per Annum ohtnined Ity an 
Engine vorking with Bowson Gaa, — It will be nsefol to place 
alongside these estimates of the cost of a steam h.p. an estimate 


of the cost of a h.p. obtained from a gas engine. For com- 
parison the very careful estimate of Professor Witz may be 
taken based on experimental trials of an engine of 112 i.h.p. or 
77 effective h.p. worked with Dowson gas. The total cost of 
the engine with pump and pipes was 944/. or 8*2/. per i,h.p. 
The gas generator cost 280Z., or 25/. per i.h.p. Foundations and 
erection (without buildings) cost 68/., or 0*61/. per i.h.p. The 
total cost without buildings was therefore 11*3/. per indicated or 
17*2/. per effective h.p., a cost about equal to that of a steam 
engine plant of the .same power. Professor Witz takes the 
cost of anthracite at 25^. a ton and coke at 28^. a ton. He 
allows for interest and depreciation 15 per cent. The gas 
consumption is taken at 84 c. ft. per effective h.p. hour, which 
allows nothing for irregular working. Professor Witz's figures 
are reduced to a year of 3,000 working hours. 

Table XVII.— Cost op an Effective H.P. in a Gas Engine using 
Dowson Gas, per Tear of 3,000 Working Hoxtbs 


Interest and depreciation at 15 percent 2*78 

Anthracite and coke 2*36 

Petty stores -40 

Wages -96 


The cost appears to be slightly less than, that of a steam 
engine of corresponding power. The cost is equivalent to O'bld. 
per effective h.p. hour. 

Cost of Steam Power in Central Stations, — The case of a 
central station worked by steam power differs from those 
previously considered, in consequence of the excess of plant 
required and the waste due to working against a varying load. 

In such a generating station, whether supplying electricity 
or energy in any other form, it is usually necessary to work night 
and day. Part of the engines must work 8,760 hours in the 
year, but for a large fraction of the time much of the plant is 
standing idle. The demand for motive power purposes is 
greatest during the day hours, that for lighting during the 
evening hours ; during part of the night the demand for any 
purpose is very small. It follows that the plant required must 
be much larger than that which would be required to meet the 
average demand, if that could be supplied uniformly during the 


24 hours. Further, there must be a reserve of power, so 
that any engine or boiler can be laid aside for examination or 
repair without hindering the work of the station. That reserve 
will be taken to be 25 per cent, of the whole power. 

The earning power of the plant depends on the average 
demand and average rate of working. The coal and labour 
depend on this also, but are increased in consequence of the 
uneconomical conditions of working. The interest and de- 
preciation must be calculated on the maximum output of which 
the plant is capable. 

How enormously the cost per h.p. per annum may be in- 
creased by the conditions which arise in central station working 
will be shown by an examination of the returns of cost at some 
electric lighting stations. 

Cost of Engines^ Boilers^ aiid Buildings recko^iedper Indicaied 
H.P.\at Full Load. — ^The following data will be assumed in the 
following calculations of the cost of central station working : — 

Table XVIII.— Cost Assumed fob Steam Plant 

Cost ill £ per i.h.pi 
at full load 
Condensing compound engine, with the necessary 
pipes, pumps, ice, for supplying condensing 

water 7*6 

Erection of engines 0-3 

Foundations • • 1*1 

— 89 
Boilers, steam pipes and auxiliary engines . .4*3 
Erection of boilers 0*5 

- - 4-8 
Buildings : — 

Engine house, boiler house, and coal store . 50 

Chimney 10 

- 60 

Total . . . 19-7 

High speed condensing engines may be obtained, erected, 
but exclusive of boilers and buildings, as low as 51. per Lh.p. 
On the other hand, there are large engines costing SOL per 
i.h.p. The cost varies with the piston speed. 

Working Cod of Efigines reckoned also ai; per LR.P. at Full 
Load, — ^The cost of coal will be taken at 7s. per ton. The cost of 
oil and petty stores will be taken at 0*28 pounds per i.h.p. per year. 
Labour is a difficult item to estimate, because it depends so 


much on management and conditions of working. * The cost of 
labour will be taken at 2Z. per i.h.p. actually exerted per 

The folio wmg rates will be assumed for interest and de- 
preciation : — 

T ^ ^ . . Per cent. 

Interest on capital cost of plant . ♦ . . 4 

Maintenance and depreciation : 

Buildings 2 

Machinery * 74 

From what has been said it will be seen that the annual cost 
of a h.p. depends on the distribution throughout the day of 
the work to be done. If the work is regular and the engine 
works at nearly full load the cost of the h.p. is comparatively 
small. On the other hand, if the work is very irregular, 
larger engines are required, the working is inefficient, and the 
cost is comparatively large. Two limiting cases will be con- 

Case h— Conditions similar to those of an Engine pumping 
to Reservoirs, night and day, all the yea/r round.— Sxxch an 
engine may be taken to work 90 per cent, of the whole year, or 
7,884 hours in the year. For one effective h.p. of work done 
there must be exerted l/0-85=M76 i.h.p., allowiug for engine 
friction. And for every M76 i.h.p., engines of 1-47 i.h.p. 
must be provided, to allow the necessary reserve. 

Case 2.— Engines working in Conditions similar to those of 
an Electric Lighting Station.— The engines work all through the 
year, but the maximum demand is four times the average 
demand. For every effective h.p. the engines must exert (neg- 
lecting the variations of mechanical efficiency) M76 i.h.p., and 
for every 1-176 i.h.p. of average demand there must be provided 
engines capable of exerting 4-70 i.h.p. during hours of maximum 
demand. Further, to allow a reserve the engine power in the 
station must be 587 i.h.p. for every effective h.p. of average 

Case 1.— Engines tvorking on a very Regular Load, in Condi- 

iions similar to those of an Engine pumping to a Reservoir. Here 

for one effective h.p. exerted during 7,884 hours annually, 
engines of 1-47 i.h.p. must be provided. Such engines may be 
taken to use 14 lbs. of steam per i.h.p. hour in test trials. But 


in ordinary work 7^ per cent, more must be allowed for leakage, 
working auxiliary engines, and less careful attention. This 
makes the consumption 15 lbs. of steam per i.h.p., or 15 x 1-176 
= 18 lbs. per effective h.p. hour. At 9 lbs. of steam per lb. 
of coal, allowing also 5 per cent, for lighting and banking fires, 
the engine would use 2*1 lbs. of coal per effective h.p. hour. 
There are very few engines working with quite so low a con- 

Table XIX.— CofcT op Installation p«b E.H.P. 

Cost of engines for 1 effective h.p., or with reserve 1-47 

i.h.p. « 1-47 X 8-9 13*3 

Cost of boilers - 1 -47x48 70 

Cost of buildings « 1-47 x 6-0 8*8 

Totril 29-1 

Annual Cost op Working per Efpectivb Hobse Power 


Interest at 4 per cent. • 1*164 

Maintenance and depreciation : — 

Buildings at 2 per cent 0*176 

Machinery at 74 per cent 1-622 

Total fixed annual cost . . . 2*862 

Coal, 21 lbs, for 7,884 hours, at 7#. per ton . . . 2*687 

Petty stores 0-327 

Driving, stoking and cleaning 2-334 

Total annual cost . . .8*110 

This is equivalent to 0-24rf. per effective h.p. hour. 

Case 2. — Engines xcorTdng with very Variable Ijoad in Con- 
ditimis similur to tliose of an Electric Lighting Station. — Here, for 
one effective h.p. supplied on the average throughout the year, 
engines of 5*87 i.h.p. have to be provided. On account of 
the inefficiency and waste, due to the variation of the load, it is 
best to estimate the steam and coal from experience in similar 
cases. Probably no electric lighting station at present works 
with quite so low a consumption as 6 lbs. of coal per hour per 
electrical unit supplied. A consumption of 9 lbs. is probably 
much more common in the best managed stations. Six lbs. 
of coal per electrical unit corresponds to 3*8 lbs. per effective 
h.p. hour. 


Table XX.— Cost op Installation. 

Cost of engines for one average effective h.p., with reserve 

6-87 i.h.p. = 6-87 x 89 . 
Cost of boilers » 5-87 x 4-8 
Cost of buildings - 5 87 x fiO . 

Total cost 

62 2 


Annual Cost op Working pbb Eppbctivb H.P. 


Interest at 4 per cent 4*64 

Maintenance and depreciation : — 

Machinery at 7} per cent 6*05 

Buildings at 2 per cent 0*70 

Total fixed annual cost . . .11-39 

Coal, 3-8 lbs. for 8,760 hours, at 7ir. per ton . . . 6-20 

Petty stores 033 

Driving, stoking and cleaning 2*33 

Total annual cost .... 1 9*25 

This is equivalent to 0*5 Ic?. per effective h.p. exerted on the 
average throughout the year. 

Cost of a H.P, at exhsthig Electric Lighting Stations, — It is 
perhaps not entirely fair to take the cost of working of electric 
lighting stations as a guide to the cost of steam power. They 
have been recently established, they work under difficult con- 
ditions, and the best methods of economising cost have probably 
not yet been arrived at. On the other hand, they are central 
stations of the kind discussed in these lectures, and accounts of 
the cost of working are published in returns made to the Board 
of Trade. 

To be as fair as possible to electrical engineers the case of 
Bradford may be taken, where according to the returns a unit 
of electricity supplied is generated more cheaply than at any 
other station. In dealing with the figures in the returns, the 
charges under the heading * salaries of manager, engineer, <S:c.,' 
and those under the heading ' redemption fund/ are discarded. 
Further, half the cost under the headings * depreciation ' and 
'repairs and maintenance' is also subtracted, because under 
these headings are included charges not belonging to the cost of 
generating power. It would not make much difference if a 
larger or smaller fraction had been subtracted. After making 
these deductions, the cost of a unit of electricity supplied at 


Bradford, mainly, if not entirely, attributable to the cost of pro- 
ducing power, is 2-ld. Now, th& mechanical value of an electric 
unit is 1*34 h.p. hours. Taking the average efficiency of the 
dynamo at 0*85, then one unit corresponds to 1 •34/0-85 = 1-57 
eflfective h.p. of the engine. Calculated on this basis it appears 
that the cost of an effective h.p. per year of 8,760 hours at 
Bradford is 491. The cost of coal and petty stores alone, ex- 
clusive of all charges for labour, interest, and depreciation, 
is 14*6Z. At most other stations for which returns are made, 
the cost reckoned in the same way is considerably greater. 




In moat applications of the energy derived from fuel the 
flactoation of the demand for power involves increase of cost 
in two ways : (1) The generator of energy must be large 
enough to meet the maximum demand, so that its cost for a 
given output is greater than it would be if the demand for 
power were constant. (2) In the working of such a large . 
generator there is waste of fuel and increased cost of super- 
intendence. In steam central stations which must supply 
energy throughout the twenty-four hours of the day the de- 
mand for power fluctuates very greatly, and the increased 
cost of power due to this is serious. Probably electrical 
central stations for lighting are those in which the fluctua- 
tion is greatest of all, and hence electrical engineers more 
than others have sought means of storing energy in times of 
small demand, to be used in times of large demand. 

To completely meet a varying demand by generators of 
energy worked at a uniform rate, there must be an amount 
of storage of energy satisfying two conditions. Twenty-four 
hours may be taken as the natural period of the fluctuation 
of demand for energy. In each twenty-four hours there will 
usually be two periods, one in which the demand falls below 
the average demand, and one in which the demand exceeds 
the average demand. The excess of energy to be supplied 
during one period will, on the average, be equal to the de- 
ficiency in the other period. If the generators are worked 
at a uniform rate, then all the energy supplied in excess of 
the mean demand in one period must be taken from storage, 
and must have been put into store during the period in 
which the demand fell below the mean demand. But this is 

r 2 


not the only condition to be satisfied. With some kinds of 
storage the rate at which energy can be taken out of store 
is unlimited. In other cases it is limited, and then the 
storage must be so arranged that the rate at which energy 
can be taken out of store is equal to the difference between 
the maximum rate at which energy is required and the mean 

Fly-wheel Storage. — All steam engines are provided with 
fly-wheels which store and re-store part of the energy generated 
in the cylinders. Let w be the weight of a fly-wheel in 
tons, V its velocity of rim in feet per second ; then the 
total kinetic energy is (approximately enough for the present 
purpose) 2,240 w (y^l2g). For a range of variation of velocity 
from V, to Vj the amount of energy alternately stored and re- 
stored is 

2,240 w^>"^ foot lbs. 

Suppose a fly-wheel weighs twenty tons, its rim velocity 
cannot generally exceed fifty feet per second, for reasons of 
strength. If a 5 per cent, variation of speed is permitted (which 
is more than is usually allowed), the amount of energy 
alternately stored and re-stored will be 169,800 foot lbs., 
or 0086 h.p. hour, an insignificant quantity compared with 
the work which would be done by an engine with such a 
fly-wheel. If, however, the fluctuation of demand for enei^ 
occurred in half a minute, the fly-wheel would supply in that 
time 10- 2 8 h.p., which might have a very useful effect in 
diminishing variation of speed. The function of a fly-wheel 
is therefore to meet the fluctuations of demand for energy in 
very short intervals of time, and it has no sensible effect in 
regulating the variation of demand and supply over longer 

Gasholder Storage, — The distribution of gas is not 
strictly a distribution of energy, but only of the means of 
conveniently obtaining it. But a gas-lighting distribution 
is analogous to a distribution of energy, and the demand 
varies nearly as much as in an electrical distribution. The 
gas engineer is happy in having a convenient and cheap 
means of storage. Usually about twenty-four hours' supply 
of gas is stored in the gasholders at a gas-generating station. 


Hence the gas-making plant can be worked at an almost 
uniform rate day and night. Taking 25 c. ft. of gas 
S3 capable of yielding one effective h.p. hour of energy, it 
appears that gasholders cost about 5s. 6d, per effective h.p. 
hour stored. Mr. Trewby puts the cost of gasholders at a 
London station at 10,000Z. per million cubic feet of gas sup- 
plied per day. Taking these gasholders to contain twenty-four 
hours' supply, and reckoning thirty cubic feet of gas per h.p. 
hour, the gasholders cost only about 6s. per h.p. hour of storage 

A station supplying one million cubic feet per day, con- 
sidered as a power station, works virtually at 1,666 average 
effective h.p. during the whole twenty-four hours. The cost 
of the gasholders adequate to meet any fluctuation of demand 
comes to only 61, per average effective h.p. supplied. Allowing 
10 per cent, for interest and depreciation, the storage adds about 
1 28, per effective h.p. to the annual cost of the power. 

Hydraulic Storage. — Hydraulic storage will be discussed in 
another chapter. It is sufficient here to state that in 
hydraulic systems energy is stored in two ways, by hydraulic 
accumulators and by reservoirs. In the accumulators the 
total amount of energy stored is so small that it is only 
sufficient to meet momentary fluctuations of demand. The 
limitation is due to the great cost of accumulators, which 
amounts to something like 300Z. per h.p. hour of storage 
capacity. The accumulator is like the fly-wheel of an engine. 
By pumping water to an elevated reservoir very large 
amounts of energy may be stcwred at not very great cost if 
local conditions are favourable. The pumped water descending 
again will re-store the energy by working hydraulic motors. A 
reservoir on an hydraulic system, like a gasholder on a gas 
supply, may be made large enough to completely meet all 
fluctuations of demand, so that the pumps can be worked at a 
uniform rate throughout the twenty-four hours. 

Compressed Air St(yrage, — In systems for distributing energy 
by compressed air there is always more or less storage of 
energy, partly in special receivers, partly in the system of 
mains. The volume of compressed air, by expanding, gives up 
energy independently of a supply from the compressing plant. 
There is a fall of pressure, and there must be some limit, due to 


this fall, at which further energy cannot usefully be drawn from 
store, but must be supplied by the compressors. 

For the purpose of diminishing fluctua^tions of pressure, 
small reservoirs of a capacity about equal to the air supply in 
three to five minutes are suflScient. For storage of energy 
much larger reservoirs are required. A very large air reservoir 
(400,000 c. ft. capacity) was at one time projected for the Paris 
compressed air system, but it was never carried out. It was in- 
tended that this should fill with water under a head of 260 ft. 
as the air was drawn oflT, the water being again driven out when 
the air supply from the compressors exceeded the demand. In 
that case the pressure would remain constant. More commonly 
the reservoirs merely supply part of the air they contain by 
expansion with diminishing pressure. 

In the Portsmouth Dockyard compressed air system, for 
instance, there are eight air receivers of a total capacity of 
18,000 c. ft. The normal (gauge) pressure is 60 lbs. per sq. in. 
Let p^ = 75 lbs. (absolute) ; p^ = 15 lbs. ; Vj = 18,000. Then 
the work to fill the reservoirs, assuming isothermal compression 
and neglecting friction, would be — 

144 _pi Vj loge(pJp^ 
= 311,100,000 foot lbs., 

or 157 h.p. hours. This agrees suflSciently with a statement by 
Mr. Corner that the receivers can be filled by the 200 i.h.p. 
compressing plant in one hour. Not all this work can be 
recovered by calling on the store in the receivers. Suppose the 
pressure can be reduced to pj = 40 lbs. per sq. in. by gauge, or 
55 lbs. absolute, before it is too low to work the motors. Then 
the work recovered would not exceed — 

144 V, {p^ loge (p,/p.)— JP2 log. (pJPn)} 
= 125,600,000 foot lbs., 

or 63 h.p. hours. Part of this would of course be lost in 
inefficient action of motors. Mr. Comer states that about half 
the machines on the air mains can be worked for about two 
hours with air drawn from the receivers, the compressors being 
stopped. This means probably that they are driven at their 
ordinary intermittent rate of working for two hours. It is 
obvious that such an amount of storage as this, though it does 


not equalise supply and demand over the twenty-four hours, 
may have an important effect in regularising the working of the 
compressors, engines and boilers, and may not only be a con- 
venience in permitting stoppage for slight repairs and in other 
ways, but may greatly reduce the waste of fuel and steam due 
to variation of demand for power. 

Accumulator cw Battery Storage, — The electrical engineer 
would be glad to have a means of storage equivalent to a gas- 
holder. For a time it was thought that such an equivalent had 
been found in the storage battery. The use of such batteries is 
limited to continuous current systems, and they have besides the 
practical defects — (1) that the maximum rate of discharge is 
limited, and (2) that about one-fifth of the energy stored is wasted. 
Nevertheless they would have been an extremely important factor 
in electric central station working but for their excessive cost. 
With a twenty-four hour load-line, such as that of most electric 
lighting stations, the amount of storage required to enable the 
generators to work at a uniform rate may be defined thus. The 
battery must be capable of supplying energy at a rate equal to 
three times the mean rate of supply for the twenty-four hours. 
Also it must be capable of storing during one part of the twenty- 
four hours, and re-storing in the other, about half the whole 
supply for the twenty- four hours. The cost of storage batteries 
prohibits their employment on this scale in large stations. 
Employed in a limited way, they serve some useful ends. In 
some stations they supply the energy required for ten to thirteen 
hours out of the twenty-four, during which time the engines 
are stopped. They diminish the fluctuation of load of the 
engines during the time in which they are running, storing 
energy not required in the external circuit. But they do not 
obviate the necessity for having a varying number of engines at 
work. Professor Kennedy puts the case well when he says 
that they ' enable the station to be shut down for some hours 
and act as fly-wheels, smoothing the irregularities of supply.' 
The accumulator battery, however, is inferior to the fly-wheel in 
the rate at which it will absorb and give out energy to meet 
momentary fluctuations of demand. 

Cost of Accumulator Batteries. — From data given me by Pro- 
fessor Ayrton it appears that eight Epstein cells tested in the 
laboratory would work at oneh.p. and store a charge for two and 


a half h.p. hours. The cells cost, without allowance for buildings^ 
insulation or switching arrangements, or for waste of energy, 
20Z. That is, the bare cost of the cells amounts to 20Z. per h.p. 
reckoned on their maximum rate of working, or to 81, per h.p. 
hour stored. Suppose a station working at an average of 500 h.p. 
The maximum demand in the twenty-four hours would be 2,000 
h.p., of which 1,500 would have to be supplied from the battery. 
The cost of the battery to supply energy at the necessary rate 
would be 30,000Z. During twenty-four hours the quantity of 
energy supplied would be 12,000 h.p. hours, half of which must 
be stored. Batteries of sufficient capacity would cost 48,000L 
Here the latter condition determines the cost. Taking interest 
at 5 per cent, and maintenance and depreciation at 12^ per 
cent., the annual cost of the battery would be 8,400Z., or nearly 
171, per h.p. of average rate of working of the station. This 
is the bare cost of the cells, without buildings, adjuncts or 

In a project for lighting Frankfort-on-Main, Mr. Oskar von 
Miller and Mr. Lindley provided large secondary battery 
stations. The batteries had a capacity of 11,700 ampere hours, 
and were capable of supplying a current of 3,500 amperes 
at 100 volts. The batteries with wood platforms, insulation, 
&c., were taken to cost 25,100Z., and the buildings for them 
11,600Z. This is equivalent to a capital cost of 23Z. per h.p. 
hour of storage capacity, or 78Z. per h.p. reckoned on the 
assumed maximum rate of working. The difference between 
this and the previous calculations is that it includes necessary 
adjuncts, buildings, and reserve of storage to meet contin- 

Thei-mal Storage, — Secondary batteries being too costly as a 
means of storage, except on a very limited scale, the question 
arises, Is any other means of storage available in conjunction 
with steam engines ? Some means of hydraulic storage will be 
considered later : such means are rarely applicable for the storage 
of steam power. Lately Mr. Druitt Halpin has proposed a 
system of thermal storage which appears in many respects to 
meet the conditions required. 

Energy is first obtained in steam power stations in the 
form of heat. Can the heat be directly stored ? Heat is a 
very unprisonable form of energy, escaping through all bodies 


and in all directions. But in New York steam is transmitted 
through miles of pipes, and by reasonable jacketing the loss 
of heat is reduced to a moderate percentage of that carried. 
In a properly constructed storehouse for heat, with reservoirs 
closely packed and presenting little external surface, the 
radiation loss need not be large. 

For storage, heat must be imparted to a material body of 
large heat capacity. It is easily given to water in boilers of 
ordinary construction. A body of water, highly heated in a 
well insulated chamber, will store a large quantity of heat. 
To permit the water to be heated it must be kept under 
the pressure corresponding to its temperature. Water heated 
above 212° and kept from vaporizing by pressure may be 
conveniently termed superlieated water. The task of storing 
a mass of heated water presents no mechanical or physical 

It is a condition of any system of heat storage for 
central stations that the energy stored should be recoverable 
whenever and at any rate of supply required. Superheated water 
fulfils the condition. If the pressure is reduced, steam is 
generated instantly and in controllable amount. The steam 
generated can be used in the engines to produce mechanical 
energy as it is wanted. 

Mr. Halpin's plan is therefore to communicate heat in 
boilers to a body of water. The heated water is stored in 
reservoirs under pressure. From the reservoirs steam is taken 
through a pressure-reducing valve exactly when and in what 
quantity it is required. Mr. Halpin proposes that the heat 
reservoirs should be under a pressure of 265 lbs. per sq. in. 
(absolute) when fully charged, the corresponding temperature 
being 406** F. He proposes that the steam engines should be 
worked at 130 lbs. per sq. in., corresponding to 347'' F. The 
total heat stored when the reservoirs are fully charged is the 
diflTerence of the total heat of the water at 406° and at 347° F., 
or the heat due to a range of temperature of 59°. Every 
pound of water falling in temperature through that range will 
yield 61 thermal units of heat. But the total heat required to 
generate a pound of steam at 130 lbs. per sq. in. from water at 
347° is 868-8 Th. U. Consequently 14^ lbs. of water falling 
in temperature from 407** to 347° will yield a pound of steam. 


To allow for radiation loss and imperfect working, this may 
be taken at 16 lbs. of water per ponnd of steam. A 
simple cylindrical reservoir 8 feet in diameter and 30 feet long 
will contain 84,000 lbs. of heated water. Such a reservoir 
would be capable of generating under the conditions supposed 
5,250 lbs. of steam at 130 lbs. per sq. in. 

The steam consumption may be taken to be, per effective 
h.p., 18 lbs. per hour in condensing and 25 lbs. in non-con- 
densing engines. Hence one such reservoir would store 286 
effective h.p. hours if the steam is used in condensing engines, or 
210 effective h.p. hours if the steam is used in non-condensing 

If the reservoir were fully charged and discharged daily it 
would yield 104,400 and 76,660 effective h.p. hours of stored 
energy yearly in the two cases. 

A reservoir 30 ft. by 8 ft. would cost, erected with ample 
allowance for buildings and appendages, 4702. -As it is not ex- 
posed to fire its deterioration would not be considerable, and 10- 
per cent, would be sufficient to cover interest, maintenance, and 
depreciation. Hence the first cost of such reservoirs reckoned 
on their storage capacity, and the annual cost per h.p. added to 
stored energy by the cost of storage, would be as follows : — 

Cost of reaenroin per j Annnal ooet of storage 
effectiye h.p. hour of , per h.p. sappUed 
storage capacity from reservoirs 

I £ 

CondeDsiDg plant . i 1*64 

Non-oondensing plant I 2-24 


The cost in the last column is the cost due to storage ot 
8,760 effective h.p. hours annually. 

Mr. Halpin's plans appear to be practicable and to promise 
considerable economy in stations where the load fluctuate& 
greatly ; but they are untried, and it would not be fair to omit to- 
point out that there are details of working involving difficulties 
which must be met. On the plan shown the steam must be 
generated in the tanks only, and a perfect circulation must be 
secured, perhaps by putting the storage tanks in series or groups. 
To prevent steam being generated where it is not wanted, and 
where it would be embarrassing, the temperature must not rise 



in any part of the system above the temperature due to the 
steam pressure in the steam space. Hence a large volume of 
circulation must be maintained. Other ways of working are 
however possible, if the method described proves to involve too 
much difiBculty. 

The cost on the mean annual h.p. supplied is not incon- 
siderable, but it is not prohibitive. The waste in irregularly 
working stations is so large that, yrimd fade, it may be assiuned 
that there is economy in storage on Mr. Halpin's system. But it 
must be remembered that this system attacks the boiler waste 

thsrmal ttoraol boiliim ano tanka 
Fig. 17. 

only, and leaves the engine waste due to varying load untouched. 
To a certain extent the latter losses can be mitigated by sub- 
division of the engines. 

Arrangement of Thermal Storage Beservoirs on Mr, Halpin^s 
Plan. — It is possible that the best way of working thermal 
storage tanks is not yet known, but one arrangement proposed 
is shown in fig. 17. The steam boiler A is completely filled with 
water, the storage tank B nearly so. The two are in free com- 
munication by a system of circulating pipes. There is an 


ordinary feed pump supplying water direct to the boiler or the 
storage tank. But instead of keeping the water in the boiler at 
a nearly constant level, the level in the storage tank is kept 
nearly constant. In addition there is a circulating pump 
maintaining a rapid current of water from the boiler to the 
storage tank, and consequently back from the storage tank 
to the boiler. Water heated in the boiler is constantly being 
sent to the storage tank, and water cooled by disengagement 
of steam is returning to the boiler. The steam spaces of the 
tanks are all in communication. The pressure there will be the 
steam pressure due to the hottest tank. The steam required 
is taken off through a reducing valve. It will then be 
generally dry or slightly superheated, in consequence of wire- 
drawing, which is advantageous for the efficiency of the 

In a station with thermal storage tanks, the boilers would 
be of a size sufficient to supply the mean demand for steam 
on the day during the year when the demand is greatest. 
The boilers would be worked continuously at a nearly uniform 
rate, like a bank of gas retorts. The heat not required in 
hours of small demand would be stored in the tanks. The 
excess of heat required to generate steam in hours of great 
demand would be taken from the store in the tanks. 

Goinpcunson of Steam Central Staiicms worked in the ordinary 
icay and worked with Thermal Storage, — The following example 
of a central station, with a load-line like that of the Ken- 
sington Electric Lighting Station, has been worked out by Mr. 
Halpin. Fig. 18 shows the ordinary midwinter load-line of 
such a station, the ordinates representing effective h.p. The 
total output in twenty-four hours is 15,600 h.p. hours, so 
that the mean rate of working is 15,600/24, or 650 effective 
h.p. For practically seven hours the demand exceeds the 
mean demand, and during that period altogether 10,450 h.p. 
hours must be supplied, which is 5,900 h.p. hours in excess 
of the 7x650 = 4,550 h.p. hours which correspond to the 
average demand in the twenty-four hours. The maximum de- 
mand for a short period is at the rate of 2,400 h.p., or 3-7 
times the mean rate. 

If this station is worked in the ordinary way, without any 
storage of energy, it must have engines and boilers able to 



develop 2,400 h.p. These will be worked to their full capacity 
for a short time only, and during seventeen hours they will have 
to develop less than 650 h.p. 

In a station with adequate thermal storage tanks, the 
engines would have to be of 2,400 h.p., but boilers of 650 
h.p. only would be necessary. The thermal storage tanks would 
supply during seven hours the excess energy, amounting 
to 5,900 h.p. hours, which would have been carried into 
them during the seventeen hours of small demand. Taking 
16 lbs. of water to supply 1 lb. of steam, 360 lbs. of water 
stored would supply one effective h.p. hour. Then the thermal 
storage tanks would have to contain 900 tons of superheated 

12 miomT 6ah I z moo ft JTm. IzniDN'^ 


Fig. 18. 

water. Twenty-four tanks 30 ft. x 8 fb. would have suffi- 
cient capacity. 

Station without storage tanks, — Eight boilers, each with 
2,000 sq. ft. of heating surface, and two boilers in reserve. Cost 
often boilers with pipes, pumps and cost of erection, 10,300?. 

Station with thermal storage tanks, — Two similar boilers, 
and one in reserve. In addition, 24 storage tanks. Cost of 
three boilers and 24 tanks with pipes, pumps and cost of 
erection, 15,000/. 

The cost of the plant is greater with the storage tanks by 
4,700Z. Taking interest and depreciation at 10 per cent., thi& 
would correspond to an annual charge of 470Z. The extra cost 
of the storage is sixteen shillings per h.p. hour of storage 


capacity, or about nine shillings per annam per h.p. of average 
output from the station. The extra annual charge of 470Z. is 
about the value of 500 tons of coal at London price. From data 
given above it will be seen that at existing electric lighting 
stations the consumption of coal is at least 4 lbs. per effective 
]i.p. hour. This for the station under consideration would 
correspond to an annual consumption of 17,000 tons of coal. If 
the saving of waste in the thermal storage station due to work- 
ing the boilers regularly instead of irregularly amounted to only 
one half-pound per effective h.p. hour, the annual saving of coal 
would be 1,270 tons. 

Cases still more favourable for the application of the thermal 
storage system are those where heat is now absolutely thrown 
away. The destructor for ash-bin refuse, which has already 
been described, must be worked continuously day and night. 
This makes it difficult to utilise the heat. But with thermal 
storage tanks the heat might be captured and stored for use at 
hours when mechanical work had to be done. In such a case 
the advantage of thermal storage would seem to be very great. 
One other similar case has been thought of by Mr. Halpin. 
In the production of lighting gas about 312 lbs. of coke are 
burned per ton of coal carbonised, and about 6,240 lbs. of 
furnace gases escape from the retort bank at a temperature of 
1,200° F. per ton of coal carbonised. The heat in these pro- 
ducts of combustion is now entirely thrown away. If they 
were taken through the flues of a boiler, they might be reduced 
to 600® before escaping. They would furnish about 759 lbs. 
of steam per ton of coal carbonised, or about 25 effective h.p. 

The gas industry is a very extensive one, and the aggregate 
waste of heat is enormous. At present the chief difficulty in 
utilising it is that there is no continuous work requiring to be 
done. But if the heat can be stored and used when wanted the 
case is very different. 

Since these lectures were delivered Mr. Halpin has pointed 
out some modifications of his system. The plan described above 
of storing superheated water is a system of complete storage 
permitting the boilers to be worked continuously at average 
load. But by merely heating a quantity of feed water sufficient 
to supply the boilers during the period in which the load ex- 


ceeds the mean load, an amelioration of the conditions of working 
is secured. This may be termed partial thermal storage. The 
feed is heated to the temperature of the boiler steam only, and 
the difficulties which may attend the use of very high pressures 
in the storage tanks are avoided. A subordinate advantage is 
that the feed water in the storage tanks may be made to deposit 
part of its lime salts in the tanks, where it is less injurious than 
in the boiler, and more easily removed. 



Where there exists a natural waterfall with a considerable and 
regular flow, and where natural conditions are favourable for 
the construction of the necessary works, water power is generally 
much cheaper than steam power. The water costs nothing ; the 
cost of maintenance of the hydraulic machinery and the cost of 
superintendence are small. The annual cost of power consists 
almost entirely of interest charges on the capital expended in 
works and machinery. The power obtained in this way is 
regular, controllable and convenient. 

So well is this recognised that industries like those for the 
electric reduction of metals, which involve the expenditure of 
large quantities of mechanical energy, seek localities where water 
power is available as a first condition of successful operation. 
In this country there is comparatively little water power. The 
drainage areas are comparatively small, and the flow of the 
streams is irregular. Coal also is cheap and abundant. Hence 
water power is of secondary importance. But in some other 
countries water power is hardly subordinate to steam power in 
manufacturing industries. The extension of methods of trans- 
mitting power to a distance would make many natural water- 
falls available which are at present unutilised. It is because 
the transmission of power is now receiving so much attention 
that there is a remarkable revival of interest in the utilisation of 
water power. 

With few exceptions, water power has hitherto been em- 
ployed only in the immediate neighbourhood of the natural 
waterfall. Where there has been a distribution of water power 
to different factories, it has commonly been effected by conveying 
the water itself to the factories, where the power is developed 


and used. In some cases water has been conveyed for mining 
and manufacturing purposes very considerable distances. But 
a more convenient and cheaper method of transmitting power 
derived from water would greatly increase the availability of 
this source of energy, and, in some countries, would change 
the relative importance of steam power and water power. 

It appears from a report by Mr. Weissenbach that, in 
1876, 70,000 h.p., derived from waterfalls, were in use for 
manufacturing purposes in Switzerland. It is estimated that 
the total available water power in Switzerland amounts to 
582,000 h.p. Putting the annual value of a h.p. at 6Z., this 
<X)rresponds to a total annual value of 3^ million pounds. Sup- 
posing it used to replace steam power, there would be an annual 
saving of 1^ million tons of coal. It is stated that, at the 
present time, Switzerland pays annually to other countries 
800,000Z. for coal.* The greatest part of this could be saved, 
if its natural wealth of water power were rendered available. 
The recognition of the importance of water power is now exciting 
great interest in Switzerland, and many factories are either 
using water power, or making preparations to do so. 

The utilisation of water power often involves the construction 
of large permanent works, such as river dams, reservoirs, and 
canals. Mr. Emery estimates that at Lawrence, in the United 
States, 200,000Z. was spent on works, independent of the 
hydraulic machinery, and at Lowell a still larger sum.^ 

Such extensive works can be best executed by an association, 
in the interest of many consumers. Thus is created a water- 
power company, who establish what is virtually a Central Water 
Power Station, and a distribution of power at a rental to con- 
sumers. In the American cases, as already stated, the water 
itself is distributed in canals to consumers, at a level permitting 
the creation of a waterfall at the mill or factory. But in certain 
other cases a further step is taken : the Water Power Company 
utilise a natural fall, and erect the necessary turbines, and 
then transmit the power in the form of mechanical energy to 
consumers. Installations of such a kind, now of a quite 
TBspectable antiquity, were erected at Schaffhausen, Freiberg, 

* Beif er. Berechnung der Turbinen, 

« • Cost of Steam Power/ C. K. Bmery ; Trans. Am. Soe. of Electrical 
.Bn^neert, vol. x. p. 123. 


Ziiricli, and Bellegarde. In these cases the method of trans- 
mitting power adopted, admirable as it was, had limitations, and 
the extension of the works was restricted. Now that there are 
new means of transmission, the Schaffhansen and Zurich power- 
generating stations are being increased, and a new and remark- 
able installation has been erected at Geneva. 

The original project for utilising the motive power of the 
Rhone at Geneva, partly for pumping a supply of water, partly 
for motive power for industry, comprised 20 turbines of 300 h.p. 
each, or an aggregate of 6,000 h.p. Fourteen of these were at 
work in 1892. Four more of somewhat larger size will, it is 
expected, be constructed by 1898. When these are at work the 
whol» available water power in Geneva will be utilised. But 
it is foreseen that the demand for power will not then have 
been satisfied. The total receipts for the installation reached 
22,500Z. in 1891, and were increasing 2,200/. annually. The 
municipality of Geneva has determined to provide for future de- 
mands, and plans are being studied for utDising 12,000 h.p. at 
a point on the Rhone six kilometres below Geneva, whence the 
power will be distributed electrically. At Biberist, near Soleure, 
Messrs. Cuenod, Sautter, & Co. have utilised 360 h.p., and trans- 
mitted it 28 kilometres electrically. At Genoa, water power, due 
to surplus fall along a line of water main, has been utilised at 
three stations. The greater part of the energy is transmitted to 
Genoa for electric lighting and power purposes. 

These are cases where water power has been utilised and 
distributed which are actually in operation. But many other 
schemes, some of them on a still larger scale^ have recently been 
projected. On the United States side of the Falls at Niagara an 
immense work is in progress for obtaining 100,000 h.p., and 
distributing it electrically. The rock tunnel for this amount of 
power is completed. A turbine wheel pit for three 5,000 h.p. 
turbines is nearly complete, and another for turbines of 6,000 
h.p. intended to drive a paper mill. A project for utilising a 
still larger amount of power on the Canadian side is under con- 
sideration. There has been a project for utilising 10,000 h.p. 
on the Dranse at Martigny ; another for utilising -20,000 h.p. 
at a point 17 kilometres above Lyons. In Sweden there is a 
project to transmit power from the Dal River at Mansbo to the 
Norberg mining district, a distance of 10 miles. There is a pro- 



ject to transmit power from a fall on the Jadal River to Otter- 
Bund, a distance of 11 miles. There are projects to transmit 
power, from the falls of TroUhatten and the river Motala, to 
Gothenburg and Nordkoping. It is stated that works are 
actually in progress for obtaining 11,000 h.p. at Orizaba, in 
Mexico, on a fall of 115 feet. The power is to be distributed 
electrically to factories. 

Water Power in the United States of America. — It is in the 
United States of America that water power is most largely used, 
where it is in most direct competition with steam power, and 
where data for a comparison of their relative advantages can 
best be obtained. Interesting information as to the extent to 
which water power is utilised in the United States is given in a 
paper by Mr. G. F. Swain read before the American Statistical 

The money value of the water power utilised in the United 
States is very considerable. From the returns of the Tenth 
Census it appears that, in 1880, there were 55,000 water-wheels 
and turbines, of an aggregate of 1,250,000 h.p. At oL per h.p. 
per annum the water power utilised is worth 6,250,000i. a 
year. • 

The comparison of the relative amount of water and steam 
power is interesting. Taking the whole of the United States, 
36 per cent, of the power used in manufacturing was, in 
1880, water power, and 6 i per cent, steam power. In certain 
industries the proportion of water power was greater. In the 
manufacturing of cotton and woollen goods, of pjaper and of 
flour, 760,000 h.p. derived from water and 515,000 h.p. derived 
from steam were employed. In the North Atlantic division 
4-81 water h.p. are utilised per square mile. 


N. Atlantic 
S. Atlantic 
N. Central 
S. Central 

The United Slates . 

Water power jier 


Steam power per 




' ' Statistics of Water Power employed in Manufacturings in the United 
States,' by O. F. Swain ; Ameriocm Statistical AstocioMofif March 1888. 




Fig. 19 is a map, taken from Mr. Swain's paper, which shows 
that over a considerable area of the United States the water 
power used exceeds the steam power. It should, however, be 
pointed out that, in the decade 1870-80, during which the total 
power used increased 45 per cent., 9 per cent, of the increase was 
due to water power and 91 per cent, to steam power. It is 

GULF Of M£Xt€0 

Fig. 19. 

possible that, under the new conditions now obtaining, the present 
decade will show a greater relative increase of water power. 

Ameiiccm Method of Distributing Water Power, — ^The method 
in which water power is distributed in America to a number of 
consumers is almost peculiar to that country. A Water Power 
Company is formed which undertakes the construction of the 
permanent works, such as a river dam, sluices, and distributing 


canals. In New England, there are five water power Btations 
where more than 10,000 h.p. is utilised during working hours^ 
and thirteen stations where more than 2,000 h.p. is utilised. 
The water is distributed to mill-owners, who construct the 
turbines and pay a rental to the Water Power Company pro- 
portioned to the amount of water used. The earliest application 
of this system was at Paterson, New Jersey, where the Passaic 
River furnishes about 1,100 h.p. night and day.* At Lowell, 
Massachusetts, the utilisation of the water power began in 
1822. The Merrimac Biver has a fall of 35 feet, and furnishes at 
the minimum about 10,000 h.p. during the usual working 
hours. At Cohoes, in the State of New York, the Mohawk 
River has a fall of 105 feet. It could furnish about 14,000 h.p. 
during working hours, but is only partly utilised at present. 
At Manchester, New Hampshire, the Merrimac has a fall of 
52 feet and furnishes at the minimum about 10,000 h.p. during 
working hours. At Lawrence, Massachusetts, the Essex 
Company built a dam, forming a fall of 28 feet, and obtaining a 
minimum power of 10,000 h.p. during working hours. At 
Holyoke, the Hadley Falls Company built a dam, forming a fall 
of 60 feet and rendering a power of 17,000 h.p. available during 
working hours. 

To indicate the magnitude of some of these works it may 
be stated that at Lawrence the masonry river-dam is 900 feet 
long and 32 feet in height. The cost was 50,000Z. From this 
dam two canals extend down stream, one on each bank, and 
between these canals and the river are located the mills, occupy-^ 
ing the entire river front. On the north side the mills extend 
for a distance of more than a mile. The cost of the canal on 
the north side, 5,330 feet in length and 100 feet in width at 
the upper end, was 50,000Z. The canal on the south side, 2,000 
feet in length and 60 feet in width, cost 30,000Z. 

The case of the town of Holyoke, Massachusetts, may be 
described in somewhat greater detail. All the factories in the 
town are worked by water power, and the system is strictly a 
distribution of power from a common source to many consumers, 
at a rental proportional to the power used, although the power 
is actually developed in the mills by turbines belonging to the 
mill-owners. The Holyoke Water Power Company controls the 
• J. B. Francis ; Trans. Am, Sod of C. E., vol. x. p. 189. 


flow of the Connecticut River, which has a drainage area above 
the town of 8,144 sqaare miles. The first weir or dam was 
built in 1847, but it was carried away. A second dam of crib- 
work was built in 1849. In 1868, an apron was constructed to 
protect the rock immediately below the dam. Since then, Mr. 
Clemens Herschel has carried out extensive repairs of the dam * 
under conditions of singular difficulty with great success. The 
structure is now 130 feet in width, 30 feet in height, and 1,019 
feet in length. Prom above the weir, a canal supplies water to 
the highest line of mills. After driving turbines in these mills, 
the water is collected in a second canal, which is a supply canal 
to a second line of mills. The tail-races of these mills discharge 
into a third canal, firom which other mills are supplied before 
the water returns to the river at a point where the level is 60 
feet below the level above the dam. 

Nearly 15,000 h.p. are in use by day and over 8,000 h.p. 
at night, of which part are 'permanent powers' held under 
leases and subject to annual rental ; the balance are * surplus 
powers ' held by contracts subject to withdrawal at short notice. 
There are 139 turbine water-wheels in Holyoke, of which 59 
run about ten hours daily, and 80 run from Sunday midnight 
to Saturday midnight, or 144 hours per week. 

With the grant of land for a ruill there was leased the right 
to use a definite portion of the water power. A ' mill-power ' 
is defined as 38 cubic feet per second, on a fall of 20 feet, during 
sixteen hours per day. This is equivalent to about 63 effective 
h.p. on the turbine shaft. At the time when Mr. Herschel 
became engineer to the Wat/er Company, the water was used 
extravagantly. By introducing a system of testing the turbines 
before their erection at the mill, data were obtained from 
which the quantity of water used by the turbine, at any gate- 
opening for any head, could be calculated. At each mill, gauges 
are placed showing at any time the height of the fall. Observa- 
tions of the fall and gate-opening at each turbine are made by 
inspectors daily, and the quantity of water used is thus ascer- 
tained. The excess of water used above that granted by the 
mill lease is charged for as * surplus power.' 

It is an advantage to the mill-owners to have this surplus 

* * On the Work done for the Preservation of the Holyoke Dam in 1885/ by 
Clemens Herschel ; Trant. Am. Soc. of C, E^ vol. zv. p. 643. 


power at moderate cost, and the system introduced by Mn 
Herschel, by which the charge for water is made to depend on 
accarate measurement of the amount used, led to great economy 
in the use of the water and secured a large surplus power at 
the disposal of the mills. 

Observations of the gate-opening of each turbine and the 
fall between head and tail race are made at each mill, at least 
once in the day and once at night. Three inspectors are en- 
gaged exclusively in this work. From the daily observations 
the quantity of water used at each mill is calculated. Part of 
this is charged for, according to the terms of the lease, at a fixed 
rental. The balance is charged for as surplus power. In times 
of drought the amount of surplus power is restricted. 

The Holyoke Testing Flamed — The first flume at Holyoke 
was constructed by Mr. Emerson. The business of testing 
turbines became so important that the Power Company erected a 
new flume and buildings, under the direction of Mr. Herschel, in 
1882. This is placed between two of the canals of the Company 
where from 17 to 19 feet of head is available. It has offices 
and repairing shops, and is fitted with dynamometers of different 
sizes, a measuring weir with accurate hook gauges, gauges for 
measuring the head, clocks with electric signal bells and other 
appliances for experiment. 

Fig. 20 shows the general arrangement of the flume. The 
water enters through a 9-foot wrought-iron pipe A from the 
upper canal into a masonry ante-chamber, from which it is 
admitted to the chamber c by the regulating sluices GO. In 
this chamber there is a small turbine for working the repairing 
shops. I is a tail-race for this turbine. From, c the water 
passes over stop planks into the wheel chamber D, on the floor 
of which is placed the turbine to be tested. For cased wheels, 
the chamber D is empty of water, and a supply pipe is attached 
to the timber bulkhead L. The water is discharged through 
culverts n into the tail-race £, at the end of which is the measur- 
ing weir o ; R is a suspended platform over the tail-race. At P 
is a recess in which the measurements of the water level over 
the weir are made. 

The turbine to be tested being fixed in the chamber D, a 

' See a paper by R. H. Thurston, AmeriMn Inst, of Mechanieal BngvMert^ 


friction brake with water-cooled rim is fixed on its shaft. 

Attendants at the brake adjust the weights and the tension of 

the friction band. Other 
observers at the weir note 
the height of water over the 
weir. The revolutions of a 
counter are noted at minute 
intervals. A series of trials 
are made with each gate- 
opening, the load on the 
brake being varied to give 
difierent speeds. From the 
observations there can be 
plotted curves giving the 
discharge for each gate- 
opening at different speeds, 
and the efiSciency corre- 
sponding to each gate-open- 
ing and speed. It is not 
absolutely necessary that the 
turbine should be tested on 
the fall on which it is to be 
used, because the discharge 
very approximately varies as 
the square root of the head, 
and the efficiency does not 
vary much for different 
heads when the speed is pro- 
portional to the square root 
of the head. 

The measuring weir has 
a capacity of 200 c. ft. per 
sec. Any head from 4 to 17 
feet can be used. The cost 
of a test is 10 per cent, on 
the list price of the wheel, 
with a minimum charge of 
^30. In 1883 the number 
of wheels tested was 185. 


Belative Cost of Water and Steam Power in the United Staies. — 
In some cases local conditions are so favourable that water power 
can be developed at an almost nominal cost. In other cases^ 
with less favourable local conditions or from unforeseen contin- 
gencies, expenditures have been incurred which have made the 
cost of water power excessive, greater in fact than that of steam 

Mr. Swain puts the average cost of steam power in the States 
in favourable localities at 4Z. per h.p. per annum, and that 
of water power at about 2Z. per h.p. per annum. Both these 
estimates are so low that it may be suspected that they are 
based rather on the nominal power of the plants than on the 
average actual h.p. used throughout the year. The cost of water 
power however varies greatly. Mr. Swain states that, while in 
the North West of the United States the cost for interest, depre- 
ciation and water rental is about 2Z. 28. to 21. 5«. per h.p. per 
annum, in New Jersey it is from 12Z. to 15Z. That water power 
is nsed at all at a cost so large as this proves that it has advan- 
tages of convenience, compared with steam power, which balance 
some excess of cost. 

It is somewhat difficult to arrive at a precise knowledge of 
the cost of water power in the great works in America, because 
of the gradual way in which they have been developed and the 
wadt of complete data as to the amount expended. Mr. C. E. 
Emery,* who is probably rather prepossessed in favour of steam 
power, has made the following estimate of the cost of water 
power at Lawrence. 

He puts the total cost of the structural works at Lawrence 
at 200,000Z., and the power utilised as equivalent to 13,000 
h.p. for ten hours daily throughout the year. That makes 
the cost of structural works 15'4Z. per h.p. The cost incurred 
by the mill-owners in erecting turbines, sluices, &c., he puts 
at 91. per h.p. of the turbines, or 131. per average h.p. actually 
utilised, the turbines being generally constructed to yield 
surplus power in times of emergency. This makes the total 
expenditure 28-4Z. per average h.p. utilised ten hours daily 
throughout the year. He allows 2^ per cent, for depreciation, 
1^ per cent, for repairs, 1^^ per cent, for taxes, 10 per cent, for 

' • The CJort of Steam Power/ by C. B. Emery; Trans. Am. See. o/Elertrieal 
Sngimeen, March 1893. 

^i I V E R 6 I T Y ; 


interest, and 2 per cent, for working expenses, or altogether 
17 per cent, on the capital expenditure. This makes the 
^annual cost of a h.p. at Lawrence 4'7Z. per annum, which he 
takes to be about the same cost as that of steam power, with 
economical engines, and coal at 88. to 12«. a ton. 

This estimate is for cases where only a moderate fall is 
available ; with large falls and in favourable conditions water 
power can be obtained at a much less cost. 

Cost of Water Power at Oerieva. — It appears that at Geneva 
for the first groups of turbines erected, of 840 h.p., and for the 
river works then completed, the capital cost amounted to 60Z. 
per effective h.p. The groups of turbines subsequently erected 
have cost only 191, per h.p. The mean cost, when the present 
works are completed, will amount to 27L per effective h.p. In 
this case the water costs nothing. If we allow 5 per cent, for 
depreciation, repairs, and working expenses, and 10 per cent, 
for interest on capital, the cost per h.p. per annam will only 
amount to 4Z. In the new works, below Geneva, where 12,000 
h.p. are to be utilised, it is estimated that the whole cost for 
turbines and structural works will amount to 601. per h.p. for 
the first 2,400 h.p. When the whole installation is completed, 
the capital cost will be only 271. per h.p. 


The need of storing power obtainable from a natural water- 
fall arises out of considerations which are different from those 
which apply in the case of steam power. A river flows 
day and night with an energy which varies seasonally, but not 
from hour to hour. The work to be done in a factory or 
central station varies necessarily from hoar to hour, and in the 
majority of cases there is no demand for mechanical energy 
for half the twenty-four hours. If no means are provided 
for storing the available energy, a large part flows away, and is 

There is another reason. In the case of water power nearly 
the whole cost is due to interest on capital expended on per- 
manent structures, and an allowance for depreciation. Very 
little is due to working expenses. But with steam power only 
about one-third to one-half of the whole cost is due to permanent 


charges, and two- thirds to one-half is due to wages and fuel. If 
a steam engine stops for twelve hours out of the twenty-four, 
half the coal and wages are saved. Though the cost per h.p. hour 
is increased, it is only increased by about 25 per cent. But if 
water-power machinery is stopped for half of the twenty-four 
hours, the cost of a h.p. hour is doubled. 

At some of the American water-power stations, an induce- 
ment is held out to consumers of power to work night and day, 
a lower rate being charged for power taken at night. In other 
•cases the difiBculty is met by storing the water during the night 
so that it can be used during the day. The amount of power 
available in working hours is then doubled. One of the most 
characteristic advantages of water power is that the storage of 
energy is possible by means not excessively costly or difficult. 
Further, it is the facility of storing energy in elevated reservoirs 
which, in some cases, makes it profitable to pump water to be 
afterwards used for power purposes. 

There are two distinct methods of storing energy in hydraulic 
systems — accumulator storage, and reservoir storage. 

Perhaps, on superficial consideration, it would appear very 
unlikely that it could be profitable to pump water for power 
purposes by steam pumps. There are cases where it is so. One 
of these is the system of hydraulic high-pressure transmission 
•devised by Lord Armstrong. This system is used, and can only 
be used advantageously, to work a great number of intermittently 
working machines. A single steam engine, working almost 
continuously, pumps water which actuates a great number of 
intermittently working hydraulic motors. Naturally the fluc- 
tuation of demand for power varies a great deal, and storage is 
almost essentially necessary. Perhaps it is to the invention of 
the accumulator, a means of storing the energy of pressure 
water, that the success of the system of hydraulic transmission 
is chiefly due. 

The hydraulic accumulator is simply a vertical cylinder, with 
« heavily loaded plunger, into which the water is pumped till it 
is required, and from which it is discharged by the descent of 
the plunger. 

Let A be the area of the plunger in sq. ft. ; P the total load 
on it in lbs. Then j? = p/a is the pressure at which the water 
is delivered in lbs. per sq. ft. If h is the length of stroke of the 


accomnlator planger, in ft., then Ah is the greatest quantity of 
pressure water it will store, and 

p Ah foot lbs. 
is the energy stored when it is fully charged. 

The pressure used in systems of hydraulic transmission is 
generally 750 lbs. per sq. in. Now one of the very large 
accumulators of the London Hydraulic Power Company has the 
following dimensions : diameter of plunger, 20 ins. ; stroke, 23 
feet ; at 730 lbs. per sq. in., this accumulator, large as it is, 
stores only 2-4 h.p. hours, a comparatively insignificant quantity. 
The cost of this accumulator, reckoned on the capacity for 
storing energy, must be very large indeed. What makes the 
accumulator so important is that its rate of discharge is very 
great. It would probably supply 100 h.p. for 1^ minutes. 
Hence, like the fly-wheel, the use of the accumulator is limited 
by its costliness to meeting fluctuations of demand for energy 
in short periods of time. It cannot be used to average the 
demand and supply during long periods. This must be effected 
by varying the engine power which supplies the energy. Costly^ 
like the electric secondary battery, it has an advantage over the 
latter, that the rate of discharge is unlimited. 

If a suitable elevated site can be found, then reservoirs can 
be built of very large capacity at a cost not large per cubic unit 
stored. Let A be the mean surface area of the reservoir, h the 
variation of depth of water in the reservoir, and H the mean 
height of the reservoir water level above the hydraulic motors 
supplied. Then the volume of water in the reservoir when full 
is A A cubic feet. The gross amount of energy stored, not 
allowing for loss in pipes and motors, is 

G A H /(t foot lbs. 

(g = 62*4, the weight of a cubic foot of water), or — 

AH A/31,740 h.p. hours. 

At Zurich, for instance, the storage reservoir contains 353,000 
c. ft., at an elevation of 475 feet above the motors. It therefore 
stores 5,284 h.p. hours. 

At both Geneva and Ziirich very remarkable and extensive 
systems for utilising and distributing water power have been 
for some time in successful operation. In both cases the water 
flowing out of a large lake with a comparatively small fall is. 


atilised to famish a very congiderable and valuable power. In 
both cases it has been found convenient and economical to use 
the low pressure turbines, in the rivers flowing out of the lakes, 
to pump water to a reservoir at a great elevation, and to use 
the pumped water for ordinary motor purposes. 

Data may be taken from the Geneva installation, though the 
works in both cases are similar, and both have been financially 

At Greneva, the Rh6ne, flowing out of Lake Leman, has by 
skilful arrangements been made to afford a clear fall varying 
from 5^ to 12 feet. There have been erected in the river, 
on this fall, or shortly will be erected, 18 low pressure 
turbines, giving in the aggregate over 6,000 h.p. These 
turbines are used primarily in pumping a supply of potable 
water for Geneva. But, since 1871, there has grown up in 
Geneva a system of using water from the town mains for motor 
purposes, and it is an important secondary function of the low 
pressure turbines in the river to pump water used for motor 
purposes. To begin with, it may be pointed out that the low 
pressure turbines, together with the building and permanent 
works required to create the fall, are very costly. It is desirable 
that, to reduce the permanent charges on the energy furnished, 
they should work as long hours as possible. Further, the total 
power furnished is even now insufficient for the work to be done, 
and it is necessary that water power should not flow to waste 
during the night hours. But, even for other reasons, it became 
necessary to use storage of energy, before the present exigencies 
arose. In the earlier period of the enterprise, to maintain 
constant pressure in spite of the fluctuation of demand on the 
mains, it was necessary that the turbines should be constantly 
pumping in excess of the demand. The surplus was discharged 
through a relief valve, and involved a constant waste. To meet 
fluctuations of pressure, four large air vessels were constructed 
5 ft. in diameter and 39 ft. high. With these, in conjunction 
with the relief valves, the working was smooth and successftil, 
but necessarily wasteful. 

When the electric installation was erected at Geneva and 
driven by turbines actuated by the pumped water, the necessity 
of storage become more evident. At 4 kilometres from Geneva 
a site was found at an elevation of 390 feet above the lake, where 


a reeeiToir coald be constructed. The reseiToir contains 
453,000 cubic feet, and stores therefore 5,573 gross h.p. hours 
of energy, which can be accumulated and discharged daily. 
Allowing for the loss at the motors driven by the water, 
the effective energy stored may be taken at 4,180 h.p. hours. 
The reservoir is a covered reservoir of expensive construction ; 
but its cost does not exceed 2*42. per h.p. hour of energy 
stored. It is a work requiring little maintenance, and it hardly 
adds 3 shillings per annum to the cost of a h.p. supplied. On 
the other hand the energy stored would without it have gone to 
waste, and for this a rental is obtained of 81. per annum. 

There is now in London an admirable system of hydraulic 
power distribution, perfectly adapted to the special purposes 
to which it is applied and mechanically and financially suc- 
cessful. But it is a system of limited applicability. Large as it 
is, the number of renters of power is under 2,000, and the maxi- 
mum pumping capacity of the supply stations at present erected 
is 2,600 effective h.p. 

In the comparatively small town of Geneva, with one 
eightieth of the population of London, there is a system of power 
distribution as large as, and of more varied applicability than, 
the system in London. There were in Geneva three years ago 
137 hydraulic motors aggregating 280 effective h.p. on the low 
pressure mains, and 79 motors aggregating 1,284 h.p. on the 
high pressure mains. The use of the hydraulic power was in- 
creasing at the rate of nearly 200 h.p. per annum. Lastly, the 
power in Geneva is distributed to ordinary consumers at one^ 
fifth, and for the electric lighting station at one-twelfth, the 
London price. 

Perhaps it is fair to add that in London there are 2,500 gas 
engines, which represent a considerable aggregate power vir- 
tually supplied from a central station. Nevertheless motive 
power is more generally and more cheaply distributed in Geneva 
than in London. No doubt local conditions have favoured the 
adoption of plans for distributing power in Geneva, but perhaps 
it has not yet been fully recognised in London what an advan- 
tage cheaply distributed power is. When this is better 
recognised means may be found to make motive power more 
available as a purchasable commodity. 




A MOVING stream of water has, in the most general case, velocity, 
pressure, and elevation, relatively to some level to which it is 
descending. Its total energy per pound is the sum of its velocity 
head, its pressure head, and its elevation head. If v is the 
velocity in feet per second, p the pressure in pounds per sq. 
ft., Ih the elevation above the level to which it can descend, and 
G the weight per cubic foot, the total head is — 

H = ^ + £ + A feet 
Icj G 

which is numerically equal to the total energy of the fluid per 
pound. Hydraulic motors are generally arranged to utilise one 
or other of these components of the total energy. 

(a) The elevation bead may be utilised by allowing the 
water to descend from its highest to its lowest level in the 
buckets of a water-wheel. The wheel is driven by the weight 
of the water on the loaded side of the wheel. Gravity water- 
wheels are motors of an antiquated type, which have been almost 
completely superseded by others either more efficient or less 

(fe) The pressure head may be utilised by allowing the water 
to descend in a closed pipe at a velocity of 1 to 3 feet per 
second and then to drive a piston. The pressure on the piston 
is that due to the height of the column of water between the 
highest and lowest level, less some frictional losses which are 
comparatively insignificant. 

(c) A third way of utilising the energy of water is to allow 
the water to issue in a jet with the kinetic energy due to the 
whole height of the fall. The jet is deviated by the curved 


floats of a turbine wheel. The effort driving the wheel is numeri- 
cally equal to the change of momentum per second in the jet, 
reckoned in the direction of motion. 

Hydraulic Pressure Engines, — Pressure engines with pistons 
driven by water are necessarily slow-moving, and consequently 
are cumbrous and expensive machines, except in the case where 
very great pressures have to be utilised. In two classes of con- 
ditions motors of this kind are used with advantage : (1) For 
working cranes, lifts, &c., where slow and steady motion is neces- 
sary, direct-acting water-pressure cylinders are safe and con- 
venient; they are so convenient that it is even advantageous to 
obtain pressure water by pumping, in order to work cranes and 
lifts in this way. (2) Small rotative pressure engines are often 
used where water under moderately high pressure is available, 
especially for driving machines used intermittently. Their conve- 
nience in such cases counterbalances the defect of a low average 
efficiency, and they serve also as almost perfect meters of the 
amount of water used. 

The volume of water used in a pressure engine cylinder per 
stroke is constant, whether a small or a great effort has to be 
exerted. The engine uses as much water for light loads as for 
heavy loads. Motors of this kind have often an efficiency of 80 
per cent, at full load. But with half load the efficiency would 
be less than 40 per cent. Hence with a varying load they are un- 
economical. Various devices have been tried for obviating this 
defect. For cranes, three pressure cylinders are sometimes pro- 
vided, which can be used one, two, or all three together, according 
to the weight to be lifted. Lord Armstrong invented an 
arrangement of this kind many years ago, and a similar arrange- 
ment is used in some London warehouses. A three-power 20 
cwt. crane may have cylinders suitable to raise 4, 12 or 20 cwts., 
according as one, two, or all three cylinders are in action. 

In Switzerland, small hydraulic rotative pressure engines, 
made by Schmid of Zurich, have been extensively used in systems 
of hydraulic distribution of power. Fig. 21 shows one of these 
engines. Such engines have one or two oscUlating cylinders, 
the movement of which opens and closes the cylinder ports, so 
that no separate valve gear is required. They use the same 
quantity of water per revolution whether the load is light or 
heavy. These motors are cheap and very conveuient for small 



powers. A revolation counter is attached to each motor, and 
the charge for water is based on the counter readings. The 
friction of the engine with 100 feet of water pressure is only 
5 per cent., and the efficiency at full load is 80 per cent., apart 
from frictional losses in the supply pipe. 

The revolving engine of Mr. Arthur Bigg, which is some- 
times used as a steam engine, can be equally well used as a 
water-pressure engine. This can be provided with a very in- 
genious arrangement for varying the stroke, so that the quantity 
of water used can be adjusted to the amount of work to be done. 
The engine may have, therefore, a nearly equal efficiency at full 


¥iQ. 21. 

and light loads. The clearance varies with the variation of 
stroke, but in the case of an hydraulic engine this does not 
afiect the efficiency. From its nearly perfect balance, it can be 
run at high speeds. 

Pig. 22 shows a four-cylinder Rigg Hydraulic Engine, which 
has been working a tramway in North Wales for two years, 
driven by a natural head of water, which gives a pressure of 
206 lbs. per sq. in. at the engine. The gunmetal cylinders are 
each 6^ ins. diameter, and 9 ins. stroke. It works at 90 
revolutions per minute with full load, at 110 revolutions with 
half load, and at 130 revolutions drawing an empty waggon. 




The four cylinders are pivoted on a central stnd. Their 
plungers act on four crank pins, on a revolving crank-disc. The 
length of stroke of each plunger is twice the eccentricity of the 
central stud. By varying this eccentricity by means of the 
subsidiary hydraulic cylinders, shown in fig. 22c, the stroke of 
the plungers is varied to suit the load. A port to each cylinder 
is provided in the central boss, by which it is pivoted on the 
central stud. This boss revolves against a valve face having an 
inlet port on one side, and an exhaust port on the other. The 
cylinders receive pressure water during one half-revolution 
and exhaust during the other. The cylinder faces are kept up 
against the valve face by a subsidiary hydraulic piston attached 
to one of the cylinders. The horizontal cylinders, which vary 
the position of the central stud, have a stroke of 4^ ins. on either 

I'lG. 22. 

side of the centre of the crank disc. If the stud is placed con- 
centric with the crank-disc, the driving plungers in the four 
working cylinders have no motion relatively to the cylinders, 
and the engine can do no work. If the stud is moved eccen- 
trically to the right, the engine rotates in one direction. If it 
is moved to the left, the engine revolves in the reverse direction. 
The water is supplied by a pipe to the left hand cylinder in fig. 
22c, and thence through the hollow plunger it reaches the in- 
let ports of the working cylinders. The water pressure in the 
hollow ram tends to move it to the right. But water pressure 
can also be admitted behind the larger plunger in the right 
hand cylinder. Then the plunger moves to the left. By two 
conical valves pressure is admitted to, or released from, the 
back of the larger plunger, so that it moves to any required 
position. If both conical valves are closed, the plunger is 


locked in position. The valves are worked by hand, and the 
two cylinders and plunger form an hydranlic relay. Lubrication 
is effected by an oil pump. 

A small engine of this construction with 2^ in. pistons^ 
making 500 revolutions per minute, and working with a pressure 
of 700 lbs. per sq. in., has been driving a three ton capstan at 
Millwall for the last five years, lliere seems no reason why 
these engines should not work with a high efficiency, but no 
exact experiments on their efficiency have been published. 


The turbine is now used in almost all cases, where the use of 
a pressure engine is not dictated by the nature of the work to 
be done, or, in the case of small motors for high falls, where the 
speed of rotation of a turbine would be inconvenient. In the 
design of turbines, theory and practice have been happily united, 
and every application of motors of this kind ought to be to some 
extent a special study. The use of such motors is of peculiar 
importance in connection with the distribution of power. It is 
where cheap water power utilised by me&ns of turbines is 
available that a system of distributing power can be most safely 

Water is conveyed to a turbine, from a pentrough at the top 
of the fall, in which there is usually a strainer to keep out solid 
bodies carried by the water, in a closed supply pipe. The 
turbine consists essentially of a fixed casing and a revolving 
wheel. In the fixed casing guide passages are formed, passing 
through which the water acquires a definite velocity, and from 
which it is discharged into the wheel at a definite angle with the 
circumference. The function of the revolving wheel with its 
curved passages is to receive the water without shock, and to 
deviate it gradually, so that it expends its energy in driving the 
wheel, and is discharged with only a small fraction of energy 
left into the tail-race. The angular impulse or moment of the 
pressure driving the wheel is equal to the change of moment 
of momentum per second of the water flowing through the 
wheel. Added to the essential elements, the guide passages and 
revolving wheel, there are arrangements for controlling the 
quantity of water used, and sometimes for regulating the speed. 

H 2 


According to the genei-al direction in which water flows 
through the wheel there are radial (inward or outward) flow 
turbines, and axial flow turbines. In some turbines the water 
flows partly radially, partly axially, and these may be termed 
mixed flow turbines. No essential difierence of action depends 
on the direction of flow. In this respect the choice of a turbine 
depends on considerations of secondary importance, and on 
questions of convenience of construction. 

A much more fundamental difference arises according as 
there is or is not a pressure in the clearance space between the 
guide blades and wheel. In all the older types of turbine, the 
water left the guide passages with part only of its pressure head 
converted into kinetic energy. The velocity at the inlet surface 
of tlie wheel was that due to a part only of the whole faU. 
Such turbines are termed pressure turbines. It is a condition of 
their proper operation that all the wheel passages should remain 
continuously filled with water. If the passages are alternately 
filled and emptied, the required pressure and velocity conditions 
cannot be maintained. Consequently, for a pressure turbine 
to act with full efficiency, the water must be supplied simul- 
taneously to the whole circumference of the wheel. 

The other fundamental type of turbine is one in which there 
is no pressure in the clearance space (except of course atmo- 
spheric pressure). The water issues from the guide passages 
with the full velocity due to the whole fall. Such turbines are 
termed fiction or impulse turbines. In such turbines the water 
is freely deviated by the curved vanes of the wheel, and it is best 
that the wheel passages should not be filled by the water. 
Generally air or ventilating holes are provided to prevent the 
filling of the wheel passages. In such wheels, since each particle 
of water acts independently, it is not necessary that the water 
should be supplied to the whole circumference. Impulse wheels 
have often paitial admission. Then the diameter of the wheel 
is determined independently of hydraulic considerations. The 
circumferential speed being fixed with reference to the velocity 
of the water, the diameter may be so chosen as to give any re- 
quired number of rotations per minute. 

Pressure turbines work best when drowned, so that there is 
no lost fall due to the drop from the wheel into the tail-race ; 
or they may be placed anywhere not more than 25 feet above 



the tail-race, the lower part of the fall being utilised by suction 
pipes, or what Americans call draught tubes. Impulse turbines 
cannot be drowned, for then the free deviation of the water on 
the wheel vanes is interfered with by the tail water in the 
wheel. When the tail-water level is variable, pressure turbines 
liave an advantage, because the impulse turbine must be placed 
above the highest tail-water level, and some of the fall is 

A compromise may be effected. The turbine may be de- 
signed so that there is no pressure in the cleamnce space, and 

Fig. 23. 

the wheel passages may have exactly the form of a freely deviated 
stream. . Then the turbine acts normally as an impulse turbine, 
though the wheel passages are filled. • If the wheel becomes, 
drowned, it acts as a pressure turbine with a small pressure in 
the clearance space. Such a turbine is termed a limit turbine. 

Fig. 23 shows the guide passages and wheel vanes of an 
inward flow pressure turbine. The curved dotted strip in the 
wheel is the absolute path of the water in the wheel. The water 
enters the wheel in a direction making an angle of about 10"* 
with the circumference, and is gradually deviated till it leaves 
the wheel radially. 



Fig. 24 shows similarly part of the guide 
of an axial flow impulse turbine. 

and wheel 

Fig. 24. 

Regulation of Turbines. — The regulation of turbines for a 
varying load has presented considerable difficulty. For pressure 

turbines three modes of regula- 
tion have been tried : — (1) By 
means of some form of equili- 
brium sluice in the supply pipe 
the effective head may be varied 
to suit the work to be done. 
In this mode of regulation the 
power of the turbine is dimin- 
ished, partly by diminishing the 
flow, partly by destroying some 
of the available head. It is 
therefore uneconomical. Gene- 
rally the speed of the turbine 
has to be kept constant. Then 
at light loads the turbine has 
to be run at a speed above that 
most suitable for the balance 
of the head left, after deducting 
the resistance of the sluice. 
This further diminishes the 
efficiency. (2) Part of the 
guide blade passages may be closed, and the turbine worked 
with partial admission. This is the commonest mode of regular 



tion, because it can be e£fected by simple mechanical means. 
But it is uneconomical, because the wheel passages alternately 
come opposite open and closed guide passages. There is 
alternately flow through the wheel passages, and the water in 
them is arrested. Large losses due to eddying and imperfect 
action of the water on the wheel vanes are unavoidable. Fig. 25 
shows the efficiency of a turbine with about as good a regulation 
of this kind as is possible. The tests were made at Holyoke. 

The numbers marked against each curve are the fractions of 
the total guide passage area open. It will be seen that the 
efficiency falls off rapidly as the guide passages are closed and 
the wheel acts at ^part gate,^ Further, the speed of greatest 
efficiency gets lower as the sluices are closed. Hence for a 
turbine working at constant speed the efficiency falls off faster 
than if the speed could be varied. 

Thus, if in the case shown in fig. 25 the turbine had con- 
stantly to run at a circumferential speed of 0"7\/2grH, the effi- 
ciencies would be as follows : 

Sluice opening 
Efficiency . 







These may be taken to be good results for a pressure 

(3) The only mode of regulation of a pressure turbine free from 
obvious objections is the varying of all the guide passages in 
area simultaneously. That can only be done in certain types of 
turbine. Fig. 23 shows a plan adopted first by the late Pro- 
fessor James Thomson. The guide blades are pivoted near 
their inner ends. By moving all the guide blades together, as 
indicated by the dotted guide blade, the passages are equally 
diminished in area without much altering the direction in which 
the water enters the wheel. 

In the case of impulse turbines the regulation is much easier, 
and it is for this reason that impulse turbines have been steadily 
superseding pressure turbines where economy of water is im- 
portant. Since the efficiency of impulse turbines is not affected 
by resorting to partial admission, guide blade passages may be 
closed without much reducing the efficiency. 

Professor Zeuner made tests of two Girard impulse turbines 


of 200 h.p. on 12 feet fall, on Prince Bismarck's estate at Varzin. 

The turbine was guaranteed by the makers to give 75 per cent. 

with full sluice and 70 per cent, with half sluice. The results 

of the tests showed the efficiency to be 0-795 with full sluice 

and 0*801 with sluices half closed, the speed being the same in 

both cases. 

Table of Classification of Turbines 

I. Impulse Turbines ' II. Pressure TurHnes 

(Wheel passages not filled.) (Wheel passages filled.) 

Discharge always above the tail Discharge above or below the tail 

water. ' water or into a suction pipe. 

1. With complete admission for I Always with complete admission 
faUload. ' for full load. 

2. With partial admission for 
fall load. 

Regulation. Usually by closing Regulation, (a) by throttling; (ft) 

part of the guide passages. by closing some of the guide passages ; 

' (fl) by varj-ing the area of the guide 

; passages. 

(a) Radial flow, inward. 
O) ,. „ outward. 
(y) Axial flow. 
Mixed flow turbines are pressure turbines only. 

The Efficiency of Turldnes. — The efficiency of turbines work- 
ing with full load and designed with a view to the beat results 
is generally taken for calculation at 0*75 to 0*80, but many 
turbines of good construction when tested have given at their 
best speed efficiencies ranging firom 0'80 to 0'85. The losses 
vary in different cases, and cannot be precisely assigned by any 
general equation. On the average they have about the following 
values in per cent, of the effective work of the fall : — 

Per cent. 

Shaft friction and leakage 3 to 5 

Unutilised energy 3 to 7 

Friction and shock in guide and wheel passages 10 to 15 

Total 16 to 27 . 

Choice of Type of Turbine, — The first general condition 
which determines the choice of the type of turbine is the 
variability or constancy of the tail-water level. Since impulse 
turbines will not work drowned, it is necessary to choose a 
pressure turbine or a limit turbine when the tail water level is 
liable to rise in flood. 


The next general consideration is that the regulating appa- 
ratus of impulse turbines is simpler and more efficient than that 
of pressure turbines. Hence when the water supply is variable 
and scanty the impulse turbine is generally to be preferred. 

The third general consideration is that on very high falls 
the rotational speed of pressure turbines is inconveniently great. 
In such cases a partial-admission impulse turbine is preferable^ 
because it permits an increase of the wheel diameter and conse- 
quently a decrease of the speed of rotation. 

When there is an ample supply of water at all times the 
pressure turbine is generally adopted, unless its speed is too 
great. It can be constructed more cheaply than an impulse 
turbine. On moderate falls with variable water supply and vary- 
ing tail- water level the limit turbine is better than the pressure 
turbine, because it has a higher average efficiency. For very 
high falls the impulse partial-admission turbine is generally 
best, on account of its low speed of rotation and the simplicty 
of its regulating apparatus. 

It is a popular delusion that by some lucky trick or specialty 
of design a turbine can be made better than that of other 
makers. The principles of turbine-designing are well de- 
termined, and the various patents taken out for turbines involve 
only details of manufacture, and not anything essential to 
efficient working. 

Speedr^oveiniors for Turbines, — For some time the question 
of automatically regulating the speed of turbines was not of 
great importance. With large low-pressure turbines on a fairly 
constant fall, hand regulation was sufficient. Some forms of 
turbine are naturally stable in speed. Now that high falls are 
utilised, and especially since turbines have been applied to drive 
electrical machinery, automatic regulation has become much 
more essential. 

The regulation of the water supply to a turbine is a more 
difficult problem than the regulation of the steam supply to an 
engine. Turbine sluices are heavy, and require considerable 
power to move them. In consequence of the momentum of the 
column of water in the supply pipe the regulation must be 
gradual. Ordinary governors are too feeble to act directly on 
water sluices. A relay must be used to work the sluices put in 
action by a governor. 



In the case of some turbines a pendulum governor puts in 
action one or other of two clutches which gear the sluices to 
the turbine itself. Governing in this way is never satisfactory. 
The action of the relay lags behind the action of the governor. 
Suppose part of the load thrown off. The speed increases and 
the governor engages the clutch which closes the sluices. But 
the action is slow, and a considerable excess of speed is at- 

FlG. 26. 

tained before the closing of the sluice stops the increase of 
flpeed. The closing must then still continue till the speed begins 
to fall. But when the speed begins to fall the effort of the 
turbine is no longer equal to the resistance of the driven 
machines. The sluice has been closed too much. The reverse 
action then begins, the action of the sluice being again too slow 
to check the decrease of speed till it has fallen below the normal. 
A periodic fluctuation of speed is so set up. 


A fly-wheel which absorbs or restores part of the work 
modifies this, and gives the governor relay more time to make 
an adjustment. Messrs. Rieter have used a fluid brake to 
absorb part of the excess work with a similar object. 

For the impulse turbines in Geneva Messrs. Faesch & 
Piccard have used the relay governor shown in fig. 26. The 
governor acts directly on the piston valve, ^, of a hydraulic relay 
-cylinder. In the cylinder is a piston with two unequal faces. 
On the lower and smaller face there is a constant water pressure 
tending to raise the piston if the water in the upper chamber 
is not locked by the valve. If the governor balls rise the valve 
allows water to flow from the upper chamber, and the piston 
rises. Conversely, if the governor falls the valve admits pressure 
water from the lower chamber to the upper chamber, and, the 
upper piston being the larger, the piston descends. The move- 
ment of the ram tends to close whichever port has been opened, 
so that unless the governor continues to rise the action on the 
turbine sluices is stopped. In that way the sluices are 
operated as if they were directly connected to the governor. 
For each position of the governor there is a definitely fixed corre- 
sponding position of the sluice. 

This governor answers perfectly for impulse turbines with 
small sluices, because the whole range of action of the sluice 
■can be obtained with a small alteration of height of a sensitive 
governor, corresponding to a small variation of turbine speed. 
But there will be a variation of speed corresponding to the 
range of governor height necessary to give the required move- 
ment of the sluice, which is virtually attached rigidly to it. 
In some later governors Messrs. Faesch & Piccard have intro- 
•duced a correcting adjustment. The movement of the piston 
adjusts the height of the fulcrum B, raising it when the governor 
is rising and lowering it when the governor is falling. The 
range of action of the sluice is then made independent of any 
variation of governor height. 



The origin of the first system by which power was trans- 
mitted to distances which, at least at the time, appeared consider- 
able is interesting.^ In 1850 there were at Logelbach, near 
Colmar, in Alsace, some buildings which had formed the 
manufactory of printed calicos of MM. Haussmann. This fac- 
tory, founded in 1772, had been reduced to inaction since 1841 
by the decay of its ancient industry. A plan was sought for 
restarting the factory as a weaving factory. There was only 
one steam engine, and the buildings were scattered at consider- 
able distances. To use shafting for transmitting the power 
would have involved too much expense and loss. It occurred to 
M. C. F. Him to drive one of the buildings at a distance of 86 
yards from the engine by a steel band 2 ins. wide and one 
twenty-fifth of an inch thick, on wooden pulleys 6^ feet in 
diameter and making 120 revolutions per minute. 

The band was not entirely successful. The least wind put 
it in vibration and the pulley guides tore it at the riveted joint. 
It worked however for 18 months, transmitting 12 h.p. Then 
an English engineer, Mr. A. Tregoning, suggested the use of a 
wire cable. A wire rope, furnished by Messrs Newall & Co., 
^ inch in diameter was substituted for the band. The same 
pulleys were used with a groove \ inch deep turned in the rim. 
That cable worked for many years, iron pulleys having been 
substituted for the wooden ones. A second transmission to a 
distance of 256 yards with a cable \ inch in diameter on pulleys 
9^ feet in diameter, running at 92^ revolutions, and transmitting 

* Notice tur la Tra/rumimon telodynamique, par C. F. Him; Colmar, 1862. 
Nate tur la TransmUsion tilodynamique inrentie par M. C. F. Hiirny par 
M. da Vt& Brozelles, 1869. JSrfahrwngtreMultate uber Betrieb utuI Imstand- 
kaltung det Drahiseiltreibt ; Ziegler, Winterthur, 1871. 


SO h.p., was soon erected. Supporting pulleys were used at the 
half-distance. The author saw this transmission still in use in 
the present year. 

M. Him has stated that his chief difficulty was the con- 
struction of the pulleys. It was not till he tried a pulley having 
a dovetailed groove, filled with gutta percha as a seating for the 
cable, that he felt the problem of telodynamic transmission to be 
solved. The durability of the transmission was of the first im- 
portance from the expense and trouble of replacing the rope. 
With these pulleys neither pulley nor cable suffered excessively 
from wear. 

The amount of work transmitted by a cable is proportional 
to the product of the effective tension (difference of the tensions 
in the tight and slack sides) and the speed. To transmit power 
to great distances by manageable cables, the strongest material 
must be used for the cables, and they must be run at the highest 
practicable speed. The cables were at first of iron, now they are 
generally of steel. The largest cables which it appears to be 
practicable to use are about one inch in diameter. In order that 
the bending stress may not be excessive the pulleys are of large 
diameter, 12 feet to 15 feet usually. For the throat of the pulley 
on which the rope runs gutta percha, softwood, and leather have 
been used. At present the bottom of the pulley groove is 
usually formed of strips of waste leather forced into a dovetailed 
groove. The greatest allowable speed of rope is that at which 
the centrifugal tension of the pulley rims becomes dangerous. 
One hundred feet per second has been adopted as the greatest 
practicable speed. The pulleys are placed at maximum distances 
of 800 to 500 feet apart. The weight of the rope then ensures 
BuflScient adhesion to prevent slipping, when the ropes are 
tightened so that the deflection or sag is not inconvenient. With 
these limitations, a 1 inch rope will transmit about 330 h.p. 

Soon after the erection of the transmissions at Logelbach 
M. Henri Schlumberger transmitted the power of a turbine 86 
yards to work agricultural machinery. In 1857, at Copenhagen, 
Captain Jagd transmitted 4-5 h.p. to saw-mills at a distance of 
more than 1,000 yards. In 1858, at Cornimont, in the Vosges, 
SO h.p. was transmitted 1,251 yards. In 1859, at Oberursel, 
100 h.p. was transmitted 1,076 yards; and at Emmendingen 
60 h.p. was transmitted 1,312 yards. 


In 1862 Him stated that about 400 applications of the 
telodynamic system had been constructed by Messrs. Stein & 
Co., of Mulhouse, carrying an aggregate of 4,200 h.p. over dis- 
tances amounting altogether to 80,000 yards. Some later 
transmissions have been constructed by Messrs. Escher Wyss, of 
Ziirich, and by Messrs Rieter Brothers, of Winterthur. Him 
wrote in 1862, with pardonable exultation, that<jusqu*^ present 
la force motrice 6tait localis^e, d6sormais elle sera mobilis6e.' 

OeneredDescrijAion^ cfihe St/stem of Transmission by Wire Rope 
Cable. — The cables used are stranded ropes having 6 to 12 wires 
in each strand and 6 to 10 strands in each rope (fig. 27). The 
strands are twisted on a hemp core, and usually there is a hemp 
core to each strand. The hemp makes the ropes flexible. At 
first Swedish charcoal iron was used, now the ropes are more 
commonly of steel. They are protected from oxidation by a 

Fig. 27. 

coating of boiled oil. It is necessary to keep a spare rope in 
reserve in case of accident. 

The ratio of the tensions in the tight and slack sides is about 
2 to 1. In passing round the pulleys the rope is bent, and the 
bending stress is added to the longitudinal stress. In order 
that the bending stress may not be excessive large pulleys must 
be used. Their diameter is so chosen that the bending stress 
and longitudinal stress are about equal. As the tension in the 
slack side is less than that in the tight side, greater bending 
stress may be permitted, and consequently supporting pulleys for 
the slack side may be smaller than the driving pulleys. 

When first erected the rope stretches a good deal from the 
tension compressing the hemp cores and diminishing the twist 
of the rope. The rope must then be re-spliced, which is trouble- 
some and costly. To diminish this difficulty Messrs. Rieter 
Brothers pass the rope before use between grooved rollers, which 



compress it laterally. This consolidates the rope. In 10 to- 
ld passages through the rollers it stretches from 1 to 4 per 
cent, in length, and it diminishes in diameter about 6 per cent. 
The following table gives data of some of the most extensive 
cable transmissions : — 

Table of Telodynamic Transuissioxs 



.2 ■** 




ZOrich . 

















Ft. j 
3,170 148 
— ! 177 












^.S JO'S' >o'2 

P ' ,S5 




•07 '- 


•072 8 



— — 



•072 10 



•043' 8 



•060 6 



•064 6 



•036 7 



•088 8 



•039, 6 





The total distance of transmission at Freiberg is 6,600 feet. 

Supporting Piers. — The piers supporting the pulleys are ex- 
pensive, and each pulley over which the rope passes is itself a 
source of wear and fiictional loss of power. It is desirable, 
therefore, that the spans of the cable should be as large as 
possible. Spans of 300 to 500 feet have been commonly 
adopted. For such spans the deflection of the rope is con- 
siderable, and must be provided for by giving sufficient height 
to the piers. The deflection of the slack side of the rope is 
greatest, and hence commonly the s'ack side is placed above 
the tight side. 

Arrarigement of Spans. — Fig. 28 (i), shows a single span 
with the deflections of the tight and slack sides of the rope. 
At (n) is shown a single span, with one supporting pulley for 
the slack side of the rope. For greater distances of transmission 
a single rope may still be used as at (iii), with supporting pulleys 
for both tight and slack sides of the rope. M. Ziegler, the 
engineer of Messra. Rieter Brothers, first constructed a trans- 
mission with a series of independent spans, fig. 28 (iv). Then 
the intermediate pulleys are two-grooved. Professor Rouleaux, 



of Berlin, has proposed the arrangement shown in fig. 28 (v) 
to reduce the height of the supporting piers. The spans of the 
tight side are twice as long as those of the slack side, so that 
the deflections are nearly eqaal, and the supporting pulleys for 
the tight side are twice the diameter of those for the slack side, 



rysziurxj cr 

" ftW T 




Fig. 28. 

«o that the total longitudinal and bending stress is about equal 

To give the rope a good frictional hold on the pulley, and to 
increase its durability, the pulley grooves are bottomed with 
«ome softer material than iron. Fig. 29 shows a section of 



single and doable grooved pulley rims, with the leather strips 
in a dovetailed groove, which are now generally used. The 
pulleys are of cast iron, or of cast iron with wrought-iron arms. 
As the principal loss of work in transmission is due to the 
weight of the pulleys and rope resting on the journals, the 

Fig. 29, 

UNIT d*ji 

pulleys should be made as light as possible. The leather 
packing in the grooves requires renewal every three or four 
months, and this involves a not inconsiderable cost in main- 

The pulley stations must be substantially constructed. 
They are sometimes of timber, sometimes of iron or masonry. 
Stations at which the direction of the transmission changes are 
called change stations. The change of direction has been 

Fig. 30. 

generally effected by bevil gearing. Fig. 31 shows one of the 
masonry pulley stations of the Bellegarde transmission. Fig. 
30 is a change station with bevil gearing. Eeuleaux proposes 
to change the direction of the ropes by guide pulleys, as shown 
in fig. 32. 




Fig. 31. 

Efficiency of Cable Transmission. — The two principal sources 
of waste of work in cable transmission are the friction of the 

journals supporting the 
pulleys, and the resist- 
ance to bending of the 
rope at the points where 
it comes on and off the 
pulleys. Ziegler made 
experiments at Oberursel 
on a transmission of 
seven spans transmitting 
104 h.p. The total loss 
of work at eight stations 
was 13^ h.p., or, say, 
1-7 h.p. at each station. 
When transmitting full 
'^ ^^ ^ ^g ^^^ w i u^L-^.-" r j"^^ ^^ power, the eflSciency of 

the system is remark- 
ably high. Probably for 
moderate distances the 
eflBciency is greater than 
for any other mode of 
transmission, except 
electrical transmission. 
But the waste of work 
is the same for all loads 
transmitted, so that 
when working at less 
than full power the effi- 
ciency falls off. If from 
Ziegler's experiments 
the efficiency for a single 
span (two pulley stations) 
is taken at 0967, then 
for a transmission of m 
intermediate and two 
terminal stations the effi- 
ciency, working with full 
load, is^ m+j 

Fig 32. V = 0-967 



Number of spans 
Number of pulley stations 
Sfficiencjr .... 

1 2 



2 3 



0-967 1 -961 



7 19 

8 I 10 
•874 1 -845, 

At half load, however, the loss of work for one span would 
be the same as for fnll load, so that the efficiency of one span 
would be only 0*934. Consequently the efficiency for several 
spans would be as follows : — 

Number of spans 
Number of pulley stations 
EflSciency .... 









1 • 





TIw Oberursel Tranmnissian, — Not long after the invention 
of Him's system of wire-rope transmission, and its application 
at Logelbach, an opportunity occurred for trying it on a more 
considerable scale. A cotton mill had been built at Oberursel, 
near Frankfort, to utilise the water power of Urselbach. Two 
tangential wheels were erected on a fall of 165 feet. Their 
total power varied from 64 to 150 h.p., according to the con- 
dition of the stream. In 1860 more power was required. A 
fall of 264 feet was found above the mill, but at a distance such 
that it could not have been made available by ordinary means 
of transmission. On this fall two tangential wheels were 
erected, each yielding from 40 to 104 h.p., according to the 
condition of the stream. The water from the upper turbines 
afterwards drives the lower turbines. From the upper turbines 
the power is transmitted by wire-rope cable a distance of 3,160 
feet, in seven spans of about 400 feet. 

Leather lining was first adopted for the pulleys of this trans- 
mission. The pulleys are 12^ feet in diameter, and make 114^ 
turns per minute, the rope having a speed of 73 feet per second. 
The pulleys on intermediate stations are double -grooved, and 
there is a separate rope to each span. The cables are f inch in 
diameter, having 6 strands of 6 wires 06 inch in diameter. 
The tensions at full load are about 11,400 lbs. per sq. in. due 
to bending, and 14,200 lbs. per sq. in. due to longitudinal tension, 
or 25,600 lbs. per sq. in. altogether. The inclination of the spans 
increases this tension a little. The installation was designed by 
M. Ziegler, and carried out by Messrs. Rieter Brothers, of 

I 2 


The Ochta Transmission. — In 1864, after a serious explosion, 
the gunpowder factory at Ochta near St. Petersburg was rebuilt, 
and wire-rope transmission was adopted in order to secure the 
condition that the buildings should be at a safe distance from 
each other. The motive power of the new factory was supplied 
by two turbines of 140 h.p. each, and a similar turbine in 
reserve. The buildings were erected in three lines, one through 
the principal axis of the turbine-house, another parallel to this 
at a distance of 420 feet, and a third at right angles to it, and 
passing through the shorter axis of the turbine-house. 

In each line the buildings nearest the turbine-house were 
330 feet from it. In the first line there were eight buildings, 
requiring 100 h.p., and the buildings were 164 feet apart centre 
to centre. The second line contained twelve buildings placed 
230 to 330 feet apart, and requiring 80 h.p. The third line 
contained three buildings, requiring 24 h.p. and placed about 
300 feet apart. There were therefore twenty-three buildings 
widely scattered to be supplied with power. The greatest 
lengths of transmission were 1,300, 2,300, and 2,600 feet. The 
power was transmitted by wire-rope cables, the work being 
carried out by M. Stein, of Mlilhausen. 

The Schaffhmiseih Transmissio7i. — A still more considerable 
application of telodynamic transmission, which attracted gene- 
ral attention, was made soon after at SchafThausen.' After a 
period of trade depression there was a revival of industry at 
Schaffhausen between 1840 and 1850. In the year 1850, Herr 
Heinrich Moser, of Charlottenfels, constructed a canal and 
erected the first turbine at Schafilwiusen. It then occurred 
to him that it might be possible to render useful the immense 
volume of water passing down the rapids of the Rhine in front 
of the town. An extraordinary low condition of the river, in 
the winter of 1857-8, favoured an examination of the bed of the 
river, and a Commission was appointed to mature a project. 
This Commission suggested the formation of a weir across the 
river ; the construction of a water-power station in the bed of 
the river near the left bank, where the conditions were suitable 

' Turbineanlage und Seiltrantmissian der Wasserwerkgesellschnft in Sehaff- 

hauten, von J. H. Erouauer; Winterthur, 1870. Die Wasterfen-rkgesellsckaft 

in Schaffliausen ; Schaffhausen, 1889. Fiinfundzwanzigster Jahregberiokt des 

Verwaltungsrathe* der Wassermerkgesellschaft in ScJiaffhausen, 1889; Schaif- 

iiaasen, 1890. 




for excavation ; the erection of turbines of 500 h.p. on a fall 

of 12 to 15 feet; and the transmission of the power across 

the river to the factories 

by the then new system of 

telodynamic transmission. 

The cost was estimated at 


Fig. 33 shows the gene- 
ral arrangement of the works 
at Schaffhausen, with the 
old turbine-house for the 
cable transmission and the 
new power house for electric 
transmission, which will be 
described later. 

A company was formed 
in 1864 to carry out the 
works. A weir was con- 
structed during favourable 
seasons in 1 864-6, across the 
rocky bed of the river, which "" " 
is about 500 feet wide. The 
fall immediately at the weir 
is not great, but there are 
rapids below it. By placing 
the turbine-house in the 
river-bed near the weir and 
constructing a tunnel tail- 
race 620 feet in length, a fall 
was obtained which varies 
from 15-6 to 13-7 feet. The 
turbine-house contains three 
axial flow pressure turbines 
with vertical shafts of 200, 
260, and 300 h.p., or 760 h.p. 
altogether. They gear with 
a common horizontal shaft 

by means of bevil wheels. Any turbine can be put out of gear 
by lowering the bevil wheel on the vertical shaft. 

Fig. 34 shows a section of the turbine-house, with the head 


! ! 



gates, the gearing, and the pulleys of the principal cable trans- 

About 150 h.p. is transmitted from one of the turbines to a 
factory on the hill above the turbine-house, by a steel shaft 550 
feet in length. From the same shaft also about 22 h.p. is trans- 
mitted, by a small cable passing down the left bank of the river 
and then crossing it, to a pulp factory on the right bank. This 
leaves a maximum of about 570 h.p. to be dealt with by the 
main cable transmission, which crosses the river directly from 
the turbine-house, and then passes along the right bank to the 
factories, as shown in fig. 33. 

As to the turbines, there is nothing of special interest except 
that they are constructed with a partition dividing them into 
two rings of buckets. During low conditions of the river, 
when there is a good fall, the outer ring of buckets only is 
used. When the fall is smaller both rings are used. This 
compensates a little for the variation of normal or most eco- 
nomical speed of periphery with variation of fall. The turbines 
make 34*28 revolutions per minute. The turbine regulating 
sluices are under the control of a relay governor. An auxiliary 
6 h.p. turbine works the main sluices of the inlet chamber. 
There is a friction brake on the main shaft of the turbines, 
which is thrown into action by the governor if the speed 
exceeds a certain limit. Then, if a rope breaks, the friction 
brake comes into action and stops the turbines. Connected 
with the brake is an apparatus for determining the power 
transmitted by the ropes ; but the author has not been able to 
leam whether this is satisfactorily used. 

It has been seen that the turbines drive by bevil wheels a 
horizontal shaft, each turbine being capable of being disconnected 
by putting its bevil pinion out of gear. The horizontal shaft 
makes 80 revolutions per minute, and carries at its driving end 
two principal rope pulleys of 14*75 feet in diameter, as shown in 
fig. 34. From these pulleys two cables cross the river in a single 
span of 385 feet to a pulley station in the river at the left bank^ 
where the direction of the transmission is changed by bevil gear- 
ing, and thence the transmission passes up the left bank of the 
river. The two principal rope-driving pulleys are not keyed 
on the horizontal driving-shaft, but run loose on it. Between 
them is a strong cross-head keyed on the shaft, carrying bevil 



wheels on studs, which gear into bevil wheels fixed to the driving 
pulleys. If the tension in the two driving cables is the same, 
the bevil gearing would have no action. But if there is any 




Fig. 34. 

difference of tension the bevil wheels permit one driving pulley 
to rotate faster than the other, so that equality of tension in the 
ropes is re-established. This differential gear has not so far as 


the author is aware been elsewhere adopted. Provision is made 
for keying either driving pnlley on the shaft in the event of 
one rope breaking and the power having to be transmitted 
through the other rope. The differential gear is then out of 

The gross power in the horizontal driving shaft in the 
turbine-house is about 530 h.p., or, allowing for friction, say 500 
effective h.p., to be transmitted to the factories, or 250 h.p. for 
each rope. Either rope is capable of transmitting at any rate 
a large fraction of the whole power temporarily, if the other 
rope is broken. 

This power is delivered by the ropes at the change station 
on the left bank. At that station about 22 h.p. is taken off 
by prolonging the second shaft of the bevil gearing and a sub- 
sidiary rope transmission. The remaining 478 h.p. is trans- 
mitted along the left bank to the first intermediate pulley station 
at a distance of 370 feet by a pair of cables. Thence to the 
second intermediate station, distant 345 feet, by another pair of 
cables. At 455 feet further is a second change station, at which 
the direction is again changed by gearing. Thence the ropes 
pass to two other intermediate stations. 

From the second intermediate station an underground shaft 
carries about 27 h.p. to ten small workshops, and from the 
second change station, and the third and fourth intermediate 
stations, cables are carried back across the river to factories on 
the right bank. From the first shaft at the second change 
station about 110 h.p. are distributed, partly by a special rope 
gear, partly by vertical and underground shafting, to four 
factories, one of which is the large Mosersche Gebaude ; and 
from the second shaft of this station a steel shaft transmits 
200 h.p. to Scholler's wool factory. 

Between the turbine-house and the second change station 
the cables consist of 80 wires, 0*042 inch in diameter, in 8 
strands of 10 wires each, with a hemp core. Diameter of rope, 
1*08 inch. Speed of rope, 62 feet per second. Smaller ropes 
are used in other parts of the transmission. The distribution 
has been described rather fully, because it is essential to learn 
how far wire-rope transmission can be adapted to complex con- 
ditions where many consumers require power. 

It would appear that the rather complex arrangement of 



differential gear and double cables was intended originally to 
meet the case of having to drive by a single cable, while the 
other was broken or under repair. It appears that under 
present conditions one cable would not be strong enough to 
transmit the whole power which is utilised. On the other hand, 
experience has shown that there is no necessity to duplicate the 
cable to avoid accidental stoppages. The total length of 
principal transmission is about 2,000 feet. 

The Schaffhausen installation has been an entirely successful 
undertaking, and has very greatly benefited the industries of 
the town. Some particulars of the extent to which the power 
has been utilised may be interesting : — 


Number of 

renters of 


Average total 
h.p. sapplied 

Kent from power 

_ __ 


















































































23 ! 



The charge for power, in 1887, varied from 4Z. 16s. to 6i. per 
h.p. per annum. 

The total cost of the works appears to have been reckoned at 
29,360Z. originally, and this by writing off stood at 24,664i. in 

As to the working of the system, it appears that experience 
has proved that there is a greater loss of power in transmission 
in wet and frosty weather than was originally expected. When 


the maximum power is being used there are oscillations of the 
ropes which drive the machines at irregular speed. The 
spinning factory suffered most from this, and ceased to take 
power. This has led to the construction of a new power station 
and the adoption in the new works of electrical instead of wire- 
rope transmission. 

In some excellent lectures which were delivered at the 
Society of Arts, in 1891, by Mr. Gisbert Kapp, the method of 
transmission by wire rope is compared with the method of trans- 
mission electrically, very much to the disadvantage of the 
former. * Till recently/ said Mr. Kapp, * rope transmission held 
the field absolutely, not because it was perfect, but because 
there was nothing better. Now, however, we have something 
better in electrical distribution, and the flying ropes are being 
steadily replaced by the electric conductors.' The use of the 
words * steadily replaced ' conveys a wrong impression. The 
wire ropes have not been replaced at Schaffhausen by electric 
cables, but an additional power station has been erected, 
and an electric transmission has been placed beside the rope 
transmission. It happened accidentally that, at a visit of the 
author to Schaffhausen about a year and a half since, the rope 
transmission was working while the electric transmission was 
stopped, having been temporarily disabled by a lightning 
accident. It seemed desirable to ascertain from Messrs. Rieter 
what view they took of the prospects of wire-rope transmission, 
looking to the fact that they had the opportunity of knowing 
the results of the working of rope and electric transmission side 
by side. They were good enough to send answers to some 
inquiries. They say that at Schaffhausen the rope plant is ex- 
pected to do more work than was originally provided for or in- 
tended. Also, it was the first large installation of the kind, and 
had some defects which experience has shown can be remedied. 
Electric transmission, they say, has also been found to have 
some inconveniences. They, however, do not think that electric 
transmission will compete seriously with rope transmission for 
moderate distances, such as that at Schaffhausen, as I understand 
them. On the other hand, for long distances they admit that 
electric transmission has the advantage. 

Cable TraTiSTnissioii at Fmbourg, Sivitzei'lajid, — In 1870 a 
company was formed, partly to acquire and work the forest 


owned by the town of Fribourg, partly to cany out a scheme of 
water supply, and partly to utilise and sell water power. This 
was a scheme in adyance of any previous one, because the com- 
pany acquired land, which they proposed to lease to industrial 
undertakings, including in the lease a right to a supply of 
motive power from the water-power station of the company. 

The Sarine, an aflSuent of the Aar, flows in a deep cut 
channel near the town. A masonry and concrete dam, about 
40 feet in height, was built across the river, so as to form a 
considerable storage reservoir in the river-bed above the dam. 
The turbine-house was built near the end of the dam on the 
right bank. The ravine through which the river flows is not 
suitable for sites for factories. The company therefore ac- 
quired some level land about 300 feet above the river, adjoining 
a railway and otherwise well adapted for industrial establish- 
ments. It was intended to work factories built on this land by 
power transmitted from the turbines by wire ropes. 

With the minimum flow of the Sarine, and an efiective fall 
of 35 feet, 1,700 h.p. could be obtained. Provision was made 
for two turbine-houses containing 8 turbines of 300 h.p. each. 
Only two turbines have actually been constructed, one driving 
pumping machinery for water supply, the other driving a cable 
transmission. The turbines are Girard turbines, running at 
74^ revolutions per minute. The turbine for transmission 
drives a horizontal shaft at 81 revolutions per minute by bevil 
wheels. This carries a 15-foot pulley with single groove, driving 
the cable by which power is transmitted to the plain of PeroUes. 
The principal rope transmission consists of five equal spans of 
500 feet each. The total distance to the saw-mill at which 
power is first taken is 2,500 feet, and the difference of elevation 
is 268 feet. The rope pulleys are all 15 feet in diameter, and 
the rope is 1*08 inch in diameter. The rope consists of 90 
wires -072 inch in diameter, in ten strands of nine wires with 
hemp cores. The speed of the rope is 62 feet per second. At 
the saw-mill a subsidiary transmission works a rope tramway 
incline for carrying timber, which uses 50 h.p. This has a cable 
f inch in diameter. From the shafting of the saw-mill another 
subsidiary rope transmission takes 1 20 h.p. to railway carriage 
works at a distance of 930 feet. From the carriage works there 
is a further transmission of 60 h.p. by a -^ inch rope a distance 


of 1,600 feet, and thence by ropes § inch in diameter to a 
foundry and chemical factory. The power is sold at the rate of 
8Z. per h.p. per annum. 

Cable Transmission at Bellegarde. — In 1872 a company was 
formed to utilize the water power of the Rhone at Bellegarde, 
not very far from Geneva. Phosphatic deposits occur near this 
point, and power was required in quarrying and grinding these 
minerals. It was expected also that other industries would be 
attracted to a site where power could be obtained. The Rhone 
flows between Fort de TEcluse and Seyssel in a winding gorge 
so narrow and deep at one part that in low water the river dis- 
appears. This part is termed the Perte du Rhone. A site for 
a power station was found almost in the bed of the Valserine 
near its junction with the Rhone. A tunnel was constructed 
from a point above the Perte du Rhone to the Valserine, calcu- 
lated to discharge 2,120 c. ft. per second, with a mean fall of 
36 feet. This would give nearly 7,000 h.p., but a part only has 
been utilised, and the works have not been financially as success- 
ful as was expected. 

Five Jonval pressure turbines of 630 h.p. each have been 
■erected. The power is transmitted upwards from the gorge to 
the plain of Bellegarde by wire cables, and is distributed to 
several works. There are phosphate works, a wood pulp factory 
and paper mill, a copper refinery and a pumping station. The 
power is sold at Si. to \2l. per h.p. per annum. 

The horizontal shafts driven by the turbines carry each two 
pulleys, 18 feet in diameter. The cables are 1*28 inch in 
■diameter, consisting each of 72 wires, 0088 inch in diameter, 
with a hemp core. The rope speed is 65 ft. per sec. The greatest 
span at Bellegarde is 630 feet. 

The Cable Transmission at Gokak, in India. — A large telo- 
dynamic transmission has been recently erected at Gokak, in 
the Southern Mahratta country in India. This installation has 
been carried out by Messrs. Escher Wyss, of Zttrich.* 

A river falling over a high cliff has motive power enough 
for many industries. At present three turbines of 250 h.p. each 
(750 h.p. altogether) have been erected to drive a cotton mill 
of 20,000 spindles. The water, taken at 2,300 feet above the 
fall, is led by a channel to the edge of the cliff, and thence 

* Kngineering, Jane 8» 1888. 



in a 32-iQch wrought-iroapipe. The pipe descends about 110 
feet vertically on the face of the cliff, and then is inclined at 

about 30^ to the horizontal. In the turbine-house the three 
turbines are supplied by three 2 1-inch branch pipes. The total 


fall acting at the turbines is 180^ feet. The turbine wheels are 
67 inches in diameter, and they run at 155 revolutions per 

The turbines are action or impulse turbines with partial 
admission, and they have horizontal axes, each turbine axis 
carrying a wire-rope-driving pulley. There is a sluice valve 
worked by hand, and a throttle or disc valve controlled by a 
governor to each turbine. The governor is a relay governor in 
which a belt on speed cones drives differential gearing connected 
with the throttle valve. If the governor moves the belt to 
either side of its central position, the differential gear comes into 
operation and opens or closes the throttle valve. 

The shafts of the turbines carry wire-rope pulleys 11 ^ feet 
in diameter. The rope speed is 93 feet per second. The ropes are 
1 inch in diameter. At the top of the cliff is a rope station, with 
carrier pulleys 198 and 220 feet above the turbine shafts. The 
pulley for the tight side of the belt is 1 1^ feet in diameter. That 
for the slack side is 8 feet in diameter. The distance from this 
station to the mill is 432 feet, and there is an intermediate 
carrier station for the slack side of the belt, which would other- 
wise foul the ground. There are of course 3 ropes, one to each 
turbine. The installation was set to work in October 1887. 

Advantages and Disadvantages of the Telodynamic System. — 
The telodynamic system is adapted for transmitting and dis- 
tributing power to distances of a mile or more, which are large 
compared with the distances to which power is ordinarily trans- 
mitted by shafting and similar means. On the other hand it 
cannot seriously compete with electrical transmission;- in cases 
where the distance to be covered is reckoned by many mUes. 
With this limitation it may be noted that — 

(1) It has the peculiar advantage that it transmits the 
mechanical energy developed by the prime mover directly, with- 
out any intermediate transformation. In electrical distribution 
a double transformation is necessary : a transformation into 
electrical energy by a dynamo, and retransformation back into 
mechanical energy by an electric motor. This double trans- 
formation involves waste of power and increase of capital ex- 

(2) The efficiency of transmission to such distances as those 
at Schaffhausen is undoubtedly very great. It is uncertain 


whether a similar amouDt of power could be distributed to an 
«qual number of consumers electrically with as little waste of 
energy in the process. 

(3) The telodynamic transmissions which have been at 
work, some of them since 1864, have actually worked con- 
tinuously without serious stoppage, and have only failed to 
return an adequate profit where they were undertaken on a scale 
too large for the amount of industry requiring to be supplied 
with power in the locality. 

(4) Where, as at Bellegarde and Fribourg, the power station 
is 150 or more feet below the factories driven, the telodynamic 
system has an advantage over some systems, such as the hydraulic 
system, in that there is no loss of eflSciency due to difference of 

On the other hand, it may be admitted that telodynamic 
transmissions, simple as they are mechanically, involve consider- 
able cost. The pulley piers require to be lofty and strongly 
built. The maximum length of span hitherto accomplished is 
630 feet, at Bellegarde. Experience has also shown that the 
cost of maintenance is considerable. The cables must be re- 
placed annually, and experienced workmen are necessary to make 
the long splices in the ropes. 

One distinct disadvantage of the telodynamic system is that 
no means has been found of directly measuring, by numerous or 
continuous observations, the amount of power delivered to each 
consumer. So long as the power is distributed to very few 
consumers it is possible to assess with practical fairness the 
charge to each without such measurements of the power. But 
in proportion as the consumers are more numerous, the defect 
of the system in this respect becomes more serious. 

It is in some cases at any rate a defect of the cable system 
that the amount of power which it is practically possible to 
transmit by a single cable is limited. It is not possible by 
increasing the size of the cable to transmit an indefinitely large 
amount of power. The cables become too heavy to be manage- 
able, and the pulleys too large in diameter. In the report of the 
experts advising the town of Geneva in 1889 the limit for one 
cable was placed at 100 h.p. Messrs. Eieter Brothers place it 
at 300 h.p., and that amount has in fact been transmitted ; no 
doubt the proper limit varies in different cases. 


It is also an inherent characteristic of the cable system that 
the eflBciency decreases rapidly when the distance increases 
beyond certain moderate limits. On the most favourable inter- 
pretation of the experiments the efficiency may be -96 for 100 
yards, or -93 for 500 yards : efficiencies remarkably high. But 
the efficiency falls to 060 for 5,000 yards : an efficiency by no 
means remarkably good. 




The distribution of power by pressure water was perhaps first 
suggested by Bramah, but the origination of a complete system 
of this kind is due to Lord Armstrong. Although the develop- 
ment of the system has been very gradual, and in spite of the 
fact that the eariiest hydraulic transmissions were of a very 
limited and local character, it appears that Lord Armstrong 
fiwm the first contemplated a distribution of power by means of 
pressure water, in town areas, to many consumers. The supply 
of motive-power water may be Qombined with the supply of 
water to towns for other purposes, the motors being driven by 
pressure water from the ordinary town mains. It was to such 
a system that Lord Armstrong first directed attention, and it has 
the attractive feature that no special network of mains is re- 
quired for the power water. On the other hand, the pressure is 
limited to that suitable for ordinary town water supply, and 
therefore cannot generally exceed 150 to 200 feet of head- 
Experience has shown that it is better in many cases to have a 
special system of mains for the supply of power water, and that 
it is convenient and economical in that case to use a much 
higher pressure than would be suitable for ordinary town mains* 
The small mains required for a power distribution can be made 
to carry safely a pressure impossible in the large mains of an 
ordinary town supply. For a long time the systems of hydraulic 
distribution which were constructed were of a local and limited 
character, and were high pressure systems of this kind. Only 
in a few towns, pressure water for a small number of motors was- 
obtained from the ordinary mains. In these exceptional cases, 
the price charged for the water was generally so great that it 
would have been preferable to use steam or gas engines. 

Reverting to high pressure systems, it was soon discovered 



that hydraulic transmission had great advantages for driving 
lifts, cranes, capstans, dock gates, and similar machines which 
work only for short periods and intermittently. For this 
particular purpose, it is convenient to use exceptionally high 
pressure, with small mains and comparatively small motor 
cylinders. Such high pi'essure hydraulic transmissions were 
first erected in connection with docks and arsenals. It was 
only after many years that similar systems came to be applied 
for power supply over extensive town districts. The conditions 
nnder which high pressure systems first achieved success gave 
them a special character which imposes definite limitations on 
their application. The high pressure system is almost ex- 
clusively an English system, and almost exclusively suitable for 
wording intermittent machines. For ordinary power purposes 
it is less well adapted. Comparatively recently systems ol 
hydraulic transmission at more moderate pressure have been 
carried out, which are better suited to distribute power for 
ordinary industrial purposes. 

In driving cranes and other. intermittently working machines, 
the fluctuation in the demand on the mains for pressure water 
is very great. Hence, in developing his system, Lord Armstrong 
was led very soon to consider the question of the storage of 
energy. Reservoir storage for systems in which the pressure is 
very great is not generally possible, because no site suflSciently 
elevated can be found for the reservoir. Air vessels were con- 
sidered, but the amount of energy which can be stored in that 
way is not very great, and there are practical difficulties. The 
pressure in the air vessel varies with the quantity of water in 
store, and an air pump must be used to replace the air, which is 
absorbed quickly at high pressures, and to maintain the air 
cushion. The invention of the hydraulic accumulator met the 
difficulty. The accumulator perfectly answers the purpose of 
storing such a supply of water under pressure, as is required to 
meet the momentary fluctuations of demand on pumping 
machinery, which is driving intermittently used motors. 

In an article in the * Mechanic's Magazine,' in 1840,* very 

interesting now if its date is considered, Lord Armstrong 

pointed out that when water is lifted by a pumping engine, it 

becomes the recipient of the energy expended in raising it. If 

* See the Proc. Trust, of Civil Engineers, vol. 1. p. 6G. 


the same water is used to actuate motors, it renders back the 
power conferred on it, in its descent to its original level, and 
thus becomes a medium through which the power of the pump- 
ing engine may be transmitted to a distance, and distributed in 
large or small quantities as required. Lord Armstrong showed 
that a continuously working steam pumping engine of com- 
paratively small size was capable of doing a large amount of 
distributed intermittent work, and he argued that this would be 
more economical than the employment of a number of steam 
motors to drive each separate machine. 

Soon after this Lord Armstrong invented an hydraulic crane 
of a type used ever since. The pressure water acted on a piston, 
the motion of which was multiplied by reduplicating a chain 
over pulleys. In 1845, a crane worked by pressure water from 
the town mains was erected in Newcastle, and in 1848 similar 
cranes were used by the North Eastern Railway at their Goods 
Station in Newcastle. In 1851, hydraulic transmission was 
adopted for driving cranes and working dock gates at Great 
Grimsby, at New Holland on the Humber, and by Brunei on 
the Great Western Railway. 

Ili^h and Low Pressure Systems, — Systems of hydraulic 
transmission are of two distinct types. (I) There are systems, 
which for convenience may be termed low pressure systems, with 
reservoir storage. In these the working pressure is fixed by 
local conditions, especially by conditions determining the site for 
the reservoir. Generally the pressure is not more than 400 to 
600 feet. It is the reservoir storage in these systems which 
more than anything else makes them suitable for the supply of 
power for all ordinary industrial purposes, for driving factories 
or electric light stations, for instance, involving a large con- 
tinuous demand for power, extending over considerable periods 
of time. (2) There are systems, which for convenience may be 
termed high pressure systems, with accumulator storage. The 
pressure in these systems is usually 700 to 800 lbs. per sq. in., 
or 1,600 to 1,800 feet of head. These systems, in which the 
reserve of energy is limited in amount, are most suitable for 
working cranes, lifts, hydraulic presses, and similar intermit- 
tently working machines. 

Amou7it of Energy transmitted hi Pipes by Pressure Water. — 
The velocity of water in very long pipes cannot be made great 

K 2 



without excessive frictional loss or without incurring danger 
from hydraulic shock. A velocity of 3 feet per second is very 
commonly permitted, and perhaps this might be doubled with- 
out excessive loss or risk. 

Let D be the internal diameter of the pipe in inches ; p the 
working pressure in lbs. per square inch ; H the head, in feet, 
due to the pressure, so that p = 433 H ; v the velocity in feet 
per second. Then the gross work transmitted is 

U = ^ D^pv foot lbs. per second 

= 0-34 D* H V foot lbs. per second 

or in horses-power 

h.p. = -001428 D* 2^ r 
= -000618 D«Hv. 

Gboss H.P. Transmitted bt Diffebent Mains 

Low Pressure System 
Head, 500 /f. 

1 Pressure 

Pressure System 
750 lbs. per sq. in. 

ter of main i Gross h.p. 
Q ins. transmitted 

Diameter of main 
1 in ins. 









This table is for a velocity of 3 feet per second — the velocity 
which has been ordinarily permitted. At 6 feet per second the 
power transmitted would be doubled. The effective power 
realised in fully loaded motors will be about three-quarters of 
the amount given in the table. 

At 3 feet per second and with a pressure of 500 feet, as at 
Zurich, a 12-inch main would transmit 133 h.p. and a 24- 
inch main 533 h.p. Mains of this size can be uned with such 
a pressure. With the high pressure of 750 lbs. per square 
inch, and at the same velocity as in the case of the London 
Hydraulic Power System, a 6-inch main transmits 116 h.p., and 
a 12-inch main, if it could be safely used, would transmit 
463 h.p. The horse-power is gross horse-power, without allow- 
ing for loss in the motors.* 

> The largest main hitherto used on a high pressoie system is 7} in. internal 
diameter. But see note, p. 135. 


In neither the high pressure nor the low pressure system is 
the amount of power which can be transmitted by a single main 
very great. This involves a definite limitation of hydraulic 
systems. They are best adapted for driving machines working 
only a fraction of the twenty-four hours, or for motors for small 
industries not requiring a great amount of power. 

Loss of Pressure due to Friction in the Mains, — At a velocity 
of 3 feet per second the loss of pressure per mile of main due 
to friction is about 18 lbs. per square inch in a 6-inch main; 
about 9 lbs. per square inch in a 12-inch main, and about 4^ lbs. 
per square inch in a 24-inch main. These losses are insignifi- 
cant on a high pressure system, and not very important on a 
low pressure system with distances of transmission such as are 
practically attempted. The losses of energy due to distribution 
in an hydraulic system, apart from those due to the pumping or 
motor machines, are so small in most cases that they may be 
dismissed from consideration without any very serious error. 

The loss of pressure measured in feet of head, in a main of 
length I and diameter D, in feet, at a velocity v feet per second, is 
given by the equation 

fe = A; - - 
D 2g 

where h has the following values for clean aud corroded pipes : ^ 

New and clean Old and incrusted 
D = 0-5 k = 0-024 to 0-048 

1-0 0022 0-044 

2-0 0-020 0-040 

The following table gives the loss due to friction, at a velocity 
of 3 feet per second, per mile of main : — 

Loss due to friction per mile 

Diameter of main 
in ins. 

In ft. of head i In lbs. per sq. in. 


Incrusted i Clean Incni8t«d 

6 36-5 

12 16-3 

; 24 , 7-4 

70-9 I 15-37 
32-5 1 706 

14-8 3-20 

' Machine Detifffif Unwin, Part ii. p. 7. 


30-70 I 

6-41 1 

. _ _J 



For rough calculations, the loss of pressure per mile may be 
taken at 107/d lbs. per square inch, for pipes in good order. 
The percentage loss per mile reckoned on the working pressure 
is as follows : — 

New and Clean Pipes 

in ins. 

T<oflfl per mile in per cent of total liead 






pressures in ft. of 

500 1 1,000 









With incrusted pipes the percentage loss is twice as great. 
The loss in any case is not very important for working pressures 
of more than 500 feet of head and distances of transmission 
likely to be attempted. For small working pressures or greater 
velocity in the main the frictional losses become much more 
important. That is one reason why high-working pressures are 
advantageous in hydraulic systems. 

Considerations ai-ising out of the Streiigth of the Pipes. — In 
all hydraulic systems at high pressure in this country cast-iron 
pipes have been used, with a peculiar flanged joint having two 
bolts. Mr. Ellington's experience in London shows that a main 
of this kind can be made absolutely tight and free from leakage. 
The largest mains used are 7^ inches in diameter. The working 
stress in the metal due to the water pressure is 2,800 lbs. per 
square inch. The mains are usually tested to a water pressure 
of 2,500 lbs. per square inch before laying, and to a pressure of 
800 to 1,600 lbs. per square inch after laying. 

Fig. 36 shows at A the form of joint used by Lord Armstrong, 
and at B a modification introduced by Mr. Ellington. The joint 
is made tight by a gutta-percha ring. The flanges are placed 
in a horizontal position in laying. Mr. Ellington found that 
fractures in the pipes occurred by the breaking off of one of 
the lugs for the bolts. By placing the lugs a little farther back 
on the pipe the strength was found to be greater. Probably the 
slight flexibility of the pipe line this form of joint allows is an 
important element in its success. 



The thickness of the pipes of D inches diameter, for a working 
pressure of p lbs. per sq. in,, is given by the rule — 

<= 0-000,178 D jp -h J 
Thus for p = 750 lbs. per sq. in. 

D= 4 6 7i 9 12 
t= -78 1-05 1-25 1-45 1-85 

The bolts have a stress of about 8,000 lbs. per sq. in. on the 
net section at the bottom of the thread. K d is the gross 
diameter of bolt, then approximately, for two bolts 



Thus for p = 750 

D = 4 

6 74 9 


d= 1 

li li H 


The proportional numbers in fig. 36 are for a unit = t. 
Probably solid drawn steel pipes could now be used, if a 
suitable joint for them could be devised.^ Such pipes were pro- 
posed to be used in a project submitted to the Niagara Com- 
mission by MM. Vigreux and Feray. For steel pipes a stress 


Fig. 36. 

of 15,000 lbs. per sq. in. might be allowed, and the use of such 
pipes would much extend the capability of the high pressure 
hydraulic system. On low pressure hydraulic systems ordi- 
nary socket pipes can be used. 

* Mannesman steel tubes of from 6 to 12 inches in diameter are being used 
to convey water under considerable hydraulic pressure (750 lbs. per sq. in.) at 
Antwerp. They have stamped steel flanges. 


Gonsideraticms arising out of the Weight aiid Cost of the Dis- 
trihding Maiiis. — For pipes of equal strength and at a given 
limiting velocity of flow the weight of mains is simply propor- 
tional to the horse-power transmitted. Hence, so far as cost of 
mains is concerned, the low pressure system is as economical as 
the high pressure system. Probably, however, if all practical 
exigencies are taken into account, the cost of mains is somewhat 
greater for low pressure than for high pressure systems. 

Considerations arising out of the Type of Motors driven by 
the Pressure Water. — On high pressure systems the motors used 
are almost exclusively pressure engines, that is, motors with 
reciprocating plungers or pistons. Such motors become ex- 
travagantly costly for low working pressures. The greater the 
working pressure the more conveniently and cheaply is the 
power produced by motors of this class. Hence the general 
adoption of pressures of 700 to 800 lbs. per sq. in. in high 
pressure systems. The pressure engine type of motor is 
extremely convenient for lifting machinery and hydraulic 
presses, and even for rotative motors of small size. It is not 
nearly so convenient when a large amount of power is to be de- 
veloped continuously for driving a factory. For that purpose 
turbine motors are much better, being cheaper and more easily 
regulated. It is true that some of the newer types of turbine, 
such as impulse turbines and Pelton wheels, can be used even at 
pressures of 800 lbs. per sq. in. But, on the whole, low pressure 
reservoir systems are better suited to cases where power has to 
be developed by turbines. 

The Efficiency of Hydraulic Transmission, — Very careful ex- 
periments on the efficiency of hydraulic transmission were made 
at the Marseilles Docks.* 

The dimensions of the engine and accumulator were as 
follows: — Engine. Two cylinders 21 ins. diameter, 38 ins. 
stroke. Accumulators. Bam 17 ins. diameter; pressure 51^ 
to 52^ atmospheres. The pumps were differential and double- 
acting. The following table gives some of the principal i*esults 
of the trials of the engine pumps and accumulator. They have 
been reduced from metric to English units. 

> * Docks and Walrehonses at Marseilles,' Thomas Hawthorn ; Proc, Inst. 
Civil Engineers, vol. xxiv. p. 144. 



Tb[al8 op Engine Pump and Accumulatoe at Marseilles 

stroke of 

Time of 20 

Volume of 1 
wteter ' 
forced into 
tor in eft. 1 
per sec. 


Useful 1 
stored in 

Bute of 


Slip of 


Eatioof , 

work 1 

stored to i 

1 tor in 20 

revs. In 


pimips iMjr 

li.p. of 


revs, of 
1 engine 


tor In 
ft. lbs. 
Iter sec. 

11,435 , 

in h.p. 



vork of ' 
engine 1 

' T2-40 



12 52 










•2192 , 














, 12-53 



25,850 1 













1 12-47 











33,520 1 







•3340 , 

37,090 , 


1 402 






46,420 ' 




73-4 ! 



-4710 1 

52,300 1 




72-4 , 



•5437 ' 






It appears therefore that at slow speeds of working 20 per 
cent., and at fast speeds nearly 30 per cent., of the indicated 
power is lost in engine and pump friction and the friction of the 
ram packing.^ 

The efficiency of the lifting machinery was also tested. In 
one case the water used in 1 ^ hour by five hoists, each loaded 
with 1^ tons, working simultaneously, was measured, and the 
engine was indicated at the same time. With 116 i.h.p. of the 
engine, the energy in the pressure water used amounted to 78-5 
h.p. and the useful work done in lifting to 34-6 h.p. Hence 
the useful work was 208 per cent, of the indicated work of the 
engine, or 44 per cent, of the energy supplied by the pressure 
water. In another case one hoist was used and the load varied. 
In this case the ratio of the useful work to the energy supplied 
by the pressure water varied from 15 per cent, with 1,100 lbs. 
lifted to 60 per cent, with 4,500 lbs. lifted. The cradle weighed 
1,650 lbs. in addition. 

Mr. Ellington stated, in his paper on the London Hydraulic 
Supply System, that 'the practical efficiency (brake h.p. of 
hydraulic motors) of the hydraulic system may be fixed at from 
50 to 60 per cent, of the power developed (indicated h.p.) at the 

* In the trial of the engine and pumps at Falcon Wharf, given in Mr. 
Ellington's paper, the i.h.p. was 178-5, and the pump h.p. (allowing 6 per 
cent, slip) was 139, so that the mechanical eflSciency of engines and pumps was 
78 per cent. 


central station.' * Great weight must be given to anything Mr. 
Ellington says in relation to hydraulic power supply, but it must 
be pointed out that this assumes a very high efficiency in the 
motors. If the efiSciency of engine, pump and accumulator is 
taken at 80 per cent., the highest eflSciency observed at 
Marseilles, and the pressure water is used to drive a continuously 
working, fully-loaded turbine, with an eflSciency of 75 per cent., 
the resultant eflSciency exclusive of any losses in the mains, which 
in fact are small, would be 60 per cent. But with the actual 

/^ _..- 

/ :i 








Fig. 37. 

machines used on hydraulic systems and the varying work done 
the average eflSciency must be very much lower. 

General Arrangement of a Hydraulic Transmission. — Fig. 37 
shows in a diagrammatic way the arrangement of an hydraulic 
transmission. As shown it is a closed system, but except for 
short transmissions the return main is generally omitted, and 
the motors discharge water to waste. Both reservoir and 
accumulator storage are indicated. In extensive systems more 
than one accumulator are required, and in that case they are best 
distributed throughout the district supplied. 

» * Hydraulic Power in London,' Proc. Inst. Citil En^inee^it vol. xciv. 



The highest water pressure usually adopted is 800 lbs. per 
square iDch, but a pressure nearly doable this is to be used in 
Manchester. The distributing mains must in general form a 
network, so that in case of accident any portion can be shut oft 
by stop valves without affecting the working of the rest of the 
system. The water used must be pure and free from silt, other- 
wise the valves and valve seatings and the working parts of the 
motors are injured. Sometimes the pressure water is obtained 
from the town mains. At other times water from an impure 
source is used, but it is filtered before it is pumped into the 
mains. To obviate injury from frost the pipes must be placed 
deep enough (three to four feet) underground. In some cases 
the pressure water is warmed in winter by taking the delivery 
pipe through the hot well of the engine, or by injecting steam 
into the suction well. To obviate injury from hydraulic shock, 
spring-loaded safety valves are placed on the main. Back- 
pressure valves are also used to prevent a sudden relief of 
pressure if a main bursts. These are sometimes ball valves 
hanging in a chamber, which swing so as to close the main if 
the velocity increases above a safe limit. 

Piston Motors, — For ordinary double-acting piston motors 
let d be the diameter of cylinder and s the length of stroke in 
inches ; let jp be the available water pressure in lbs. per square 
inch ; n the revolutions per minute ; H the h.p. ; 17 the eflSciency. 
V = snISGO is the mean piston speed in feet per second, and 

d^pvTf = 550 H 

d = 26*46 \/— inches 

rj may be taken at 0*7 to 0*8 ; v is usually about 2 feet per 
second ; the ratio 5/cZ is usually 1^ to 1^. 

The IIuU Hydraulic Power System.^ — This was the first scheme 
for distributing power hydraulically to many consumers. The 
principal main is 6 inches in diameter and 1,485 yards in length. 

* See Robinson, Proo, Ingt, Ciril JSnffineers, vol. xlix. 


The joints are flanged joints, with a gutta-percha ring of the 
kind generally used since. The pumping station is arranged 
for four 60 h.p. engines, of which two have been erected. Each 
engine delivers 130 gallons per minute at 700 lbs. per square 
inch pressure, corresponding to 63*6 effective h.p. There is an 
accumulator 18 inches in diameter and 20 feet stroke, loaded to 
610 lbs. per square inch. The charges were originally intended 
to be 52 Z. for one crane per annum, and less for several cranes 
in one warehouse. The charges are, however, by quantity of 
water supplied as measured by meter. The minimum charge is 
81, per machine per annum. The charge by quantity of water 
used ranges from 8Z. per annum for 1 6,000 gallons or less to 
2001. for 1,200,000 gallons, with special rates for greater 
quantities. The following short table will give an idea of the 
way the charges are graduated : — 

I Consumption of water per quarter 


Charge per l,0(iO 
gallons suppiieil 

£ 8. 




3 10 


12 10 








4,000 gallons or under 
9,000 to 10,000 
49.000 to 60,000 
99,000 to 100,000 . 
199,000 to 200,000 . 
299,000 to 300.000 . 

The charge for water taken in excess of 300,000 gallons per 
quarter is 2*5 shillings per 1,000 gallons. For 500,000 gallons 
per quarter 2*5 shillings per 1 ,000 is charged for the whole supply. 

The London Hydraulic Fower Company.^ — An Act was 
obtained in 1871 for supplying hydraulic power in London. 
The rights conferred by the Act remained dormant until resusci- 
tated by Mr. Ellington in 1882. The present company was 
constituted in 1884. In 1887, twenty-five miles of pressure 
main had been laid in London streets, and at the present time 
there are nearly sixty miles of pressure mains. These extend 
from the West India Docks and Wapping on the east to 
Kensington on the west ; from Mint Street south of the river 
to Clerkenwell and Old Street on the north. 

* The account of the London Hydraulic Power System is derived partly 
from papers by Mr. E. B. Ellington (^Proc. Tntt. C. E.^ vol. xciv. and vol. cxv.), 
partly from the reports of the Company. 


There are three principal pumping stations : one at Falcon 
Wharf, a short distance east of Blackfriars Bridge (800 i.h.p.) ; 
another at Millbank, Westminster (600 i.h.p.) ; and one at 
Wapping (1,200 i.h.p.). A fourth station in the City Boad is 
in course of erection (1,200 i.h.p.). The aggregate power of all 
the stations, when complete, will be 3,800 i.h.p., of which 
one-third is reckoned as reserve power in case of repair or 

At Falcon Wharf and Millbank all the water is taken from 
the river, but it is filtered before it is pumped into the mains. 
At Wapping part is taken from the London Dock, part from 
a well. After use by consumers it flows into the sewers. The 
power is available for use night And day all the year round. It 
is largely used for lifting machinery and for presses and pumps. 
The company claim that it can be used for electric lighting 
of particular establishments and for extinguishing fires. For 
this last purpose Mr. Greathead*s injector hydrant or hydraulic 
intensifier is applied ; a small jet of water from the high pressure 
mains is made to intensify the pressure of a larger jet drawn 
from the ordinary town mains. A fire stream is so obtained 
capable of reaching the top of high buildings without employ- 
ing a fire-engine. In 1892, there were 1,696 machines worked 
by pressure water from the company's high pressure mains, 
consuming 6,000,000 gallons per week. The quantity of water 
used by each consumer is measured by a meter on the exhaust 
pipe of the machines driven. Parkinson's meter is most used. 
Siemens' turbine meter is used to some extent, but it is in- 
accurate under the sudden fluctuations of discharge which 
occur. Kent's positive meter is also used. 

At Falcon Wharf there are four sets of compound pumping 
engines capable of indicating 200 i.h.p. They are vertical, 
with one high and two low pressure cylinders, and a pump 
plunger directly connected to each piston. At 200 feet of 
piston speed per minute each set of engines will deliver 240 
gallons per minute at 750 lbs. per square inch pressure into the 
accumulator. This corresponds to 120 effective h.p. A nine 
hours' trial of one set of engines was made in 1887, the engine 
running at constant speed and the coal used being sea-borne 
small coal. The boilers are provided with an economiser. 



Trial op Hydraulic Pumping Engines 

Total indicated horse-power 1 785 

Piston speed, ft. per min 221-4 

Steam pressure, lbs. per sq. in 82*5 

Evaporation (from and at 212°) per pound of fuel, lbs. . 10-59 

Feed water per i.h.p. hour, lbs 19-79 

Coal per i.h.p. hour, lbs 2*19 

Accumulator pressure, lbs. per ^q. in. . 750 
Effective h.p., calculated from water pumped, allowing 

5 per cent, slip 139 

Mechanical eflSciencj of engine 0'78 

Water pumped per min., gallons 265*7 

The engines consume in ordinary work 2*93 lbs. of coal per 
i.h.p. hour, which is greater than the result given above in 
consequence of the fluctuations of speed. 

Wapping Station 

UndeT^TOund Reservoirs 
Scola of F«ct 

Fig. 38. 

There are two accumulators at Falcon Wharf with rams 20 
inches in diameter and 2*3 feet stroke. Each accumulator has 
a capacity of storage equal to 2 4 h.p. hours. The filters are 
Perrett filters constructed by the Pulsometer Company. The 
filtering material is compressed sponge. The sponge is kept 
under a pressure of about 4 lbs. per square inch by a special 
hydraulic ram . It is cleansed every four to six hours by reversing 
the direction of flow and by alternately compressing and 
releasing the pressure on the sponge. 

The pumping station at Wapping,* of which a plan is shown 
in fig. 38, is a more recently constructed and larger station than 

* The Wapping Station was described in the Enyineer for January 20, 


that at Falcon Wharf. The water pumped is obtained partly 
from a well sunk into a gravel bed, partly from the London 
Dock. The pumping from the well into a tank over the boiler- 
house is effected by low lift pumps worked hydraulically by the 
pressure water. From this tank it passes through 'Torrent' 
filters constructed by the Pulsomefcer Company to underground 
reservoirs. From this it is lifted by the condenser circulating 
pumps to another tank above the boiler-house, whence it is 
pumped into the mains. The reservoir capacity is 800,000 
gallons. The engine-house contains six sets of vertical inverted 
triple expansion engines with cylinders 15 inches, 22 inches, and 
36 inches in diameter and 24 inches stroke. Each piston drives a 
single-acting plunger pump with ram 5 inches in diameter direct 
from the crosshead. The working steam pressure is 150 lbs. 
per square inch, and the hydraulic pressure 800 lbs. per square 
inch. Each set of engines will deliver 300 gallons of water per 
minute at a piston speed of 250 feet per minute. All the 
cylinders are jacketed. 

Ln a test trial the engines are stated to have worked with 
14-1 lbs. of steam and 1-27 lb. of Welsh coal per i.h.p. hour. 
The water passes from the pumps to two accumulators, with 
rams 20 inches in diameter and 23 feet stroke. One of the 
accumulators is loaded to a slightly heavier pressure than the 
other, so that one accumulator rises a little in advance of the 
other. The more heavily loaded accumulator automatically 
shuts off steam when at the top of its stroke. 

Charges for Pressure Water for Power Purposes. — The London 
Hydraulic Company make a minimum charge of 11. 5s. per 
quarter per machine. For consumers using more than 3,000 
gallons per quarter there are graduated charges of which the 
following short table gives a sample : — 

Coat of pressure , 

Gallons used per quarter 



water per 1,000 

1 £ *, 


.S,000 or under 

] 5 


10,000 . 

3 10 


50,000 . 

' 12 10 

50 ! 

100,000 . 

1 20 


200,000 . 

31 5 


300,000 . 

. 1 42 10 


2-8 i 


An excess over 300,000 gallons per quarter is charged at 
28. per 1,000 gallons. Consumers using more than 500,000 
gallons per quarter are charged 2s, per 1,000 gallons all round. 
Bates are further reduced for still larger quantities, and the 
minimum rate is l'5s. per 1,000 gallons. 

With these charges the cost of power for the kind of work 
for which an hydraulic system is best suited is small. Tlius it 
is often less than one farthing per ton lifted 50 feet. On the 
other hand, it is necessary for the purpose of this treatise to 
consider the cost of power distributed by different methods on 
some common basis. It is almost unavoidable to take the cost 
of power exerted continuously through the working day. If the 
cost of power supplied by the Hydraulic Power Company is 
reckoned, for machines working 3,000 hours per year, then the 
cost is larger than that of power obtained in other ways. The 
comparison of the cost so reckoned is instructive, although it 
may be in fairness pointed out that Mr. Ellington in his paper 
stated that he had never advocated the supply of power by the 
London Hydraulic System for continuously working engines to 
any large extent. 

To obtain one effective h.p. during 3,000 hours per annum, 
allowing an efficiency of 80 per cent, in the motor, 437,500 
gallons of water are required. Hence a consumer, taking 
50,000 gallons per quarter, would get the equivalent annually 
of 0*457 effective h.p. for 3,000 hours, and would pay for it 
at the rate of lOdl. per h.p. per annum. A consumer taking 
300,000 gallons per quarter would get the equivalent of 2*743 
effective h.p. for 3,000 hours, and would pay at the rate of 62Z. 
per effective h.p. per annum. A consumer taking 500,000 
gallons per quarter would get the equivalent of 4*573 effective 
h.p. for 3,000 hours, and would pay at the rate of 43i. 155, per 
effective h.p. per annum. It must be remembered that this is 
the cost for pressure water only, and does not include meter rent 
or interest on the cost of the motors. 

There is one other respect in which the statement just made, 
unfavourable as it is to the use of pressure water as an agent 
in distributing water power, is nevertheless too favourable. It 
follows from the incompressibility of water that nothing Tike 
expansive working is possible in a water motor. In the case 
of all reciprocating motors, and these are almost the only motors 


used with high pressure water, and with the small exception of. 
the special engine of Mr. Kigg mentioned above, it may be said 
broadly that all the motors on high pressure hydraulic systeibs 
use the same quantity of water, whether they are lightly loaded 
or fully loaded. The consequence is that not only is pressure 
water expensive as an agent for distributing power, when it is 
used as economically as is possible in fully loaded motors, but 
the cost is again increased because in practice most of the motors 
work usually at less than full load. If we take the average 
load on the motors to be not more than two-thirds the full load, 
then the cost of the power is increased 50 per cent. 

TJie Liverpool Hydraulic .Supyly System, — In Liverpool 
pressure water from the town mains was used for working 
hydraulic cranes as early as 1847. From an interesting paper 
by Mr. Joseph Parry,* it appears that the use of hydraulic power 
in this way made very slow progress. In 1877, the number of 
hydraulic machines supplied from the town mains was 89. At 
the present time there are 162 machines worked by water from 
the town mains, consuming 125,600,000 gallons per annum. 
Taking the mean pressure at 70 lbs. per sq. in., this is equiva- 
lent to 82,710 eflFective h.p. hours, or to 27 effective h.p. for 
3,000 hours in the year — a rather insignificant amount. The 
average charge for working a goods hoist is VSL per annum, 
or only lOd. per hoist per day, a small cost for the convenience 
afforded. The charge for water is 7d. per 1,000 gallons. At 
this rate the charge is equivalent to 120Z. per effective h.p. per 
year of 3,000 hours. Experiments on the quantity of water used 
by some hoists showed the cost to amount to from Gd. to lOd, 
per ton lifted 50 feet. 

There is also in Liverpool a high pressure system, which is 
to be extended. Experiments with some hoists worked on this 
system showed the cost to be from 1 jeZ. to 2^eZ. per ton lifted 
60 feet. Mr. Parry comes to the conclusion that hoists worked 
from the town mains cost more than those on the high pressure 
system, when the charge for water on the high pressure system 
does not exceed 58. per 1 ,000 gallons. 

The Birmingham Hydraulic Power System. — In Birmingham, 
as in Liverpool, water has been supplied from the town mains 

' • The Supply of Power by Pressure from the Public Mains/ Proe. Inst. 
<f Mechanical Engineers. 



• to work lifts. In 1888, there were 61 lifts and hoists thus 
worked, using 80,000 gallons per day, and yielding to the Water 
Committee of the Corporation about 1,000Z. a year. Since that 
time, a high pressure system has been carried out, which has 
the peculiarities that it belongs to the Corporation and that the 
pumping is done by gas engines.^ 

At the pumping station there are three sets of triple 
hydraulic pumps, working to a pressure of 730 lbs. per sq. in. 
These are driven by three ' Otto ' gas engines, nominally of 12, 
20, and 20 h.p., but capable of developing an aggregate of 
about 100 h.p. Ordinary lighting gas is used. The pumps 
deliver into two 6-inch mains. There are three hydraulic 
pumps, having each three plungers. For one the plungers are 
2^ ins. diameter and 9 ins. stroke. For the others the plungers 
are 3 ins. diameter and 12 ins. stroke. Each gas engine drives 
a counter-shaft by a belt, and this shaft drives the pump crank- 
shaft by gearing at 49 revs, per minute. There are two ac- 
cumulators with 20-inch rams and 20 feet stroke. A small 
* Brotherhood ' engine, worked by the pressure water, is used in 
starting the gas engines. 

Mancliester Hydraulic Power Supply, — At Manchester & 
combined scheme for supplying electricity and high pressure 
water is being carried out. A pressure of 1,600 lbs. per sq. in. 
in the hydraulic system is to be used. It is hoped that th^re 
will be economy in working the electricity and pressure water 
supply from the station. 


TJie Zwich Works. — ^The Ziirich installation is a complex 
and very interesting one.^ It was the earliest example in 
Switzerland of the application of hydraulic power, partly to 
pump a supply of potable water, partly to furnish motive power, 
from the same central station. It has grown gradually, and of 
late has been greatly extended. It comprises machinery driven 
by turbines for furnishing (a) a water supply to the town of 
Ziirich ; (h) a supply of motive power transmitted by wire rope ; 

1 See Engi'Merifigj February 12, 1892. 

« See Preller on ♦The Zurich Water Supply Power and Electric Works," 
Proc. Inst. Civil Eiigirieert, vol. cxi. 



(c) a supply of motive power transmitted by comparatively low- 
pressure water from the town mains ; (d) a supply of motive^ 
power transmitted hydraiilically, from a special reservoir at 
comparatively high pressure ; (e) an electric central station, also, 
driven by water power. 



R£9£R )^om 


Fig. .^9. 

When the works (fig. 39) were first established, the water 
supply of Zurich was obtained from a filter in the bed of the 
river Limmat near its exit from the lake. This water was 
pumped by turbines erected a little further down stream^ 
'fhere being surplus water power, a telodynamic transmission 
was erected, and part of the motive power was distributed to- 
fjEU^tories along the riverside. In 1881 the quality of the water 
was found to be inferior. After extensive investigations, it was 
decided to obtain a new supply of potable water from an intaka 

L 2 



in the lake, and to use the old water supply for motive power 
purposes only. 

The fall available in the Limniat at the pumping-station, 
and the available volume of flow, are as follows : — 

Fall, feet 

High-water level in summer 
Mean ,, ,» ,» m 
Low „ „ „ winter 




Volnme of 
water flowing 

in the river, 
c. ft. per sec. 


GroM water 
power, h.p. 


The eflTective power delivered by the turbines in the river is 
as follows : — 


For pumping filtered poteble water 237 

Supplying motive power by pressure water . , . 128 

Driving the wire rope transmission 227 

Supplying hydraulic motive power for electric lighting 

station 444 

Total 1,036 

There are two reserve steam-engines, of 300 i.h.p. each, to 
provide for a deficiency of water power. At the pumping 
station there are eight pressure (Jonval) turbines working,- up 
to from 96 to 110 h.p., according to the state of the river. 
There are also two newer turbines of about 175 h.p. each. The 
turbines have vertical shafts, and each pair drives by bcvil 
wheels a common horizontal shaft, which runs at 50 revolutions 
per minute in the case of the earlier turbines, and at 6(5 
revolutions per minute in the case of the two last erected. From 
these shafts a horizontal main shaft, 328 feet in length, and 
running at 100 revolutions per minute, is driven. To this 
main shaft any of the pumps can be coupled. The earlier 
turbines cost, with gearing, about 12Z. per h.p. ; the two larger 
turbines about 71. per h.p. There are at present in operation 
seven sets of horizontal double-acting ' Girard ' pumps. The total 
pumping capacity is 8,143,000 gallons per day. The water for 
driving the turbines is obtained by a weir in the Limmat, which 
deviates the water into a canal formed by a longitudinal em- 
bankment in the river. Sluices divide the head-race channel 
from a tail-race channel formed in a similar way. 

The pumps supply the following reservoirs : — 


Height above 



Capacity in a ft. 


To be in- 
creased to 

Low level 

Intermediate level . 
High level . . . . 
Reservoir for high pressure 
power water .... 








From the town mains water is supplied to work 180 small 
motors. The total power thus supplied is about 1 87 h.p., and 
iU cost is about 4'4:d. per h.p. hour. 

The principal supply of power, apart from that distributed 
by wire rope, is obtained from pressure water derived from the 
old Limmat filter-bed, and pumped to the special high level 
reservoir. This water is pumped chiefly during the night. The 
reservoir is about 6,000 feet from the pumping station, and is 
supplied by an 18-inch main. The effective pressure at the 
motors is about 475 feet, and the distributing mains have an 
aggregate length of 15,000 feet. 

The charge for this pressure water for power purposes varies 
from 0'6d. per h.p. hour, when at least 50,000 h.p. hours are 
taken in the year, to l'25cZ. per h.p, hour, when less than 20,000 
h.p. hours are taken in the year. For 3,000 working hours in 
the year the charge is from 7L lOs, to 16/. per h.p. per annum. 
The water supplied in this way now amounts to 42,380,000 
c. ft. per annum, yielding altogether about 900,000 h.p. hours. 
The total receipts are 1,200Z. per annum, or r08rf. per 1,000 

Besides this supply of pressure water to various consumers, 
the Electric Station (fig. 40) is ordinarily to be driven by 
pressure water from the same high-level reservoir. For this 
purpose two impulse turbines of 300 h.p. each have been erected 
for driving dynamos, and two smaller turbines of 30 h.p. driving 
exciting dynamos. Alternatively, if the supply of pressure 
water fails, the dynamos can be driven by the river turbines, or 
by the reserve steam-engines. 

The Hydraulic Worl'3 and System of Hydravlic Pover Supply 
at Geneva, — There is now in operation at Geneva one of the 




most remarkable hydraulic power stations in the world. The 
water of the river Rhone, near the point where it flows out of 
Lake Leman, is employed to drive a number of large low 
pressure turbines, giving a total of 4,500 effective h.p. These 
turbines pump pure water obtained from the lake into two 
systems of mains. The older of these, termed the low pressure 
system, the pressure at the pumps being 160 to 200 feet, is an 
extension of a previously existing system of mains used for 
supplying potable water to the town of Geneva. Although some 
of the water pumped into this system is used for power purposes, 
it is chiefly intended to supply water for domestic and municipal 
purposes. The second system of mains, termed the high pressure 
system, the pressure at the pumps being 460 feet, supplies 
potable water to some districts not reached by the low pressure 
system, but it is specially intended to afford a supply of water for 
motive power purposes to the entire area of the town. The 
demand for water, both on the low and high pressure systems, is a 
fluctuating demand, large during the day, and very small during 
the night. Hence, if the turbines in the Rhone were employed 
solely in pumping into the mains, they would not be continuously 
working, and a large part of the water power of the Rhone 
would be wasted. To meet this diflSculty an important storage 
reservoir has been constructed at Bessinges, about 4 kilometres 
from Geneva. The turbines pump water up to this reservoir at 
night, and at times when the demand for power for other pur- 
poses is insuflicient to keep them fully employed. The. energy 
derived from water flowing back from the Bessinges reservoir 
through the high pressure system represents part of the water 
power of the Rhdne, which would necessarily have been wasted 
if this means of storage had not been provided. 

The works at Geneva have gradually developed under special 
local conditions. In spite of natural and political isolation, 
manufacturing industries have for centuries flourished at Geneva. 
That they did so is partly owing to the fact that cheap water 
power could be obtained by simple forms of water-wheel placed 
in the ample and rapid Rhdne, flowing past the town. An 
industrial quarter gathered along the banks of the river, and 
factories were built even in the stream itself. As the population 
increased a water supply was required. The small aqueducts 
of spring water became insuflicient, and further recourse was 


had to the motive power of the Rhone. From the beginning of 
the eighteenth centnry, water-wheels placed in the Rhone 
pumped a water supply into the town. 

Then arose an antagonism to the utilisation of the motive 
power of the Rhdne, which for two centuries hindered the pro- 
gress of industrial enterprise at Geneva, and threatened at times 
to destroy the existing industries. The properties of riparian 
owners on the shores of Lake Leman were from time to time 
injured by the rising of the lake level. It was not unnatural 
that the landowners should attribute the disastrous inundations 
from which they suffered to the obstacles created at the outlet 
of the lake, that is to the bridges and buildings, and especially 
the factories and water-wheels in Geneva. Complaints were 
addressed by the Canton Vaud to the Federal Government at 
Berne of damage caused by the works at Geneva. Then arose 
a question of arrangements necessary to regulate the lake^level, 
and to facilitate, in time of flood, the discharge of the water. 
After 1875, the project of utilising the motive power of the 
Rhone took a new magnitude and importance, from the com- 
bination with it of plans for regulating the level of Lake Leman, 
and so ending a long and bitter controversy. 

Another local circumstance had great influence in deter- 
mining the ultimate form of the project for the utilisation of the 
motive power of the Rhone. In 1871, Colonel Turrettini,* the 
engineer under whose direction the present works have been 
constructed, had applied to the town council of Geneva to place 
small pressure engines on the mains of the then existing low 
pressure water supply. The plan of obtaining motive power in 
this way proved so successful and convenient that, in 1880, there 
were 111 motors at work, using 34 million cubic feet of water 
annually, and paying a yearly rental for power water of 2,000Z. 
The cost of the power at that time to consumers was at the rate 
of 36i. to 48i. per h.p., per year of 3,000 working hours. 

' In 1878, a private firm asked the concession of a monopoly 
of the motive power of the Rhone at Geneva, on condition of 
carrying out works necessary for facilitating the discharge from 

^ UtilUatitm det Forcet motriees du Rhone et Bfjfularisation du Lae Leman^ 
Th. Turrettlni, Ing^nienr Ck>DseiUer Administratif d416gu6 aaz travauz; 
Geneve, 1890. This admirably lUastrated memoir fully describes' the origin, 
progress, and details of all the works at Geneva. 



the lake and regulating the lake level. A similar offer was 
made in 1881. But there grew up a feeling that such works 
should be carried out by, 

and for the pro6t of, the 
town itself Finally, after 
many studies, the con- 
tract was given by the 
municipality in 1883 to 
M. Chappuis to construct 
under their direction the 
present works. 

These works have cost 
altogether 283,000Z. Of 
this sum a fraction has 
been paid by owners of 
land on the shores of the 
lake, and pai*t has been 
expended in constructing 
new sewers required in 
consequence of the altera- 
tions of river level. De- 
ducting these items, the 
cost of utilising the mo- 
tive power of the Rhone 
has already amounted to 
nearly 200,000i. 

The scheme included 
the clearing away of all 
obstacles to the free flow 
of the river, and the di- 
vision of the river by a 
longitudinal embankment 
into two portions, one 
forming a head-race to 
the turbines, fig. 41, the 
other, which was straight- 
ened and deepened, form- 
ing an outlet for the 
surplus water from the 
lake. Between the two 


divisions of the river bed are movable sluices, which keep up 
the water in the head-race channel, or discharge surplus water 
into the tail-race channel, accoi*dingto the condition of the lake. 
The scheme also included a complete reconstruction of the old 
pumping system for the low pressure water supply ; the crea- 
tion of the new system of high pressure water supply, and the 
provision of motive power by hydraulic transmission to the 
industries of the town. 

Tlie Low Pressure River Turbines, — The turbine and pump 
house is placed at the end (fig. 41) of the left-hand channel 
or head-race. The turbines are of 210 h.p. each, and 14 
groups of turbines and pumps have been erected. Four more 
groups of somewhat greater power are expected to be erected 
within the next five years. The turbines are Jonval pressure 
turbines, constructed by Messrs. Escher Wyss & Co., of Zurich. 
They have vertical shafts, and each turbine drives from a crank 
two horizontal double-acting ' Girard ' pumps, placed at right 
angles. Fig. 42 shows a cross-section of the turbine and pump 

The head at the turbines varies from oo feet, when the river 
is in flood, to 12^14 feet when the volume of flow is smallest. 
With most forms of turbine this would involve a considerable 
variation of the normal speed, or speed of greatest eflSciency. 
The turbines are skilfully arranged to meet this variation of 
head. The turbine- wheel and its corresponding system of 
guide-passages are arranged in three concentric rings. When 
the fall is great and the quantity of water used is smallest 
the outer ring only is open, and the water acts at a large radius. 
As the fall diminishes the second ring is opened and the mean 
radius at which the water acts is smaller. In the lowest 
conditions of the fall, when most water must be used, all three 
rings are open, and the mean radius at which the water acts 
is smaller still. The number of rotations of the turbine de- 
pends directly on the velocity due to the head, and inversely on 
the radius at which the water enters. Hence, as the radius 
diminishes as the head diminishes, a fairly constant speed of 
rotation is obtained. The adjustment is such that with the 
highest fall the normal speed is 27 revolutions per minute, and 
with the lowest fall 24 revolutions per minute, a variation not 
practically serious in working pumps. 



The fixed distributor over the turbine-wheel is 13*78 feet 
in external diameter and 5-74 feet in internal diameter. It is 
divided into three rings having fifty-two guide passages in the 
outer ring, forty-eight in the middle ring, and forty in the inner 
ring. The external ring has no regulating sluices, regulation 
being efiected when that ring only is open by the sluices in the 
head-race. The other rings are arranged so that over one semi- 
•circle the orifices open vertically on an annular plane surface, 
and over the other semicircle they open horizontally on a 

Scab a pMt 

Fio. 42. 

-cylindrical surface. Each ring of passages has two regulating 
sluices, one a semicircular annular plate for the orifices opening 
vertically, one a semi-cylinder for the openings which are hori- 
zontal. Each sluice can be fully o]iened without interfering 
with the openings corresponding to the other. The sluices are 
worked by gearing. The turbine-wheel is of cast iron, in two 
halves. It has wheel passages corresponding to those in the 
distributor. The effective section of flow through each turbine 
is: — 


Outer ring, fifty-two passages, each 11-02 ins. x 2'95 ias., 
giving a total of 11'75 sq. ft. 

Middle ring, forty-eight passages, each 17*72 x 2*60 ins., 
giving a total of 15*35 sq. ft. 

Inner ring, forty passages, each 17*72 x 2*36 ins.^ giving a 
total of 11*63 sq. ft. 

In low conditions of the river, when the fall is greatest 
(12*14 ft.), the turbine must discharge 211*9 c. ft. per second. 
The relative velocity of discharge is 0*65\/2y x 1214 = 18*17 
ft. per second. Then the outer ring affords sufficient area. 

In high conditions of the river, when the fall is 5*51 ft., 
the turbine must discharge 471*4 c. ft. per second. The relative 
velocity of discharge is 12*25 ft. per second. Then the three 
rings give sufficient area. 

The vertical support of each turbine consists of a fixed 
wrought-iron pillar, carrying at its top a steel step for the pivot 
and a steel revolving hollow shaft hanging from the pivot at the 
top. The pivot is 6 inches in diameter. A crank at the top of 
the shaft drives two * Girard ' double-acting pumps placed at right 
angles, from a single crank-pin. The ' Girard ' pump consists 
virtually of two plunger pumps placed end to end, the advantage 
being that the stuflBng-boxes for the plungers are accessible and 
there is no internal packing. The two pumps discharge into & 
single air-vessel placed between them. The diameter of the 
plungers of the low pressure pumps is 1*41 feet, that of the 
high pressure pumps 1*08 and 0-85 feet. The stroke is 3*61 feet,, 
and the mean velocity 188 feet per minute. The valves are 
ring valves with leather faces. The high pressure pumps supply 
mains of 20-inch diameter in one direction and of 24-inch 
and 16-inch in the other. The low pressure pumps supply two- 
mains of 20-inch diameter. 

The Hujh Pressure Reservoir at Bessirujes. — When the high 
pressure system was first put in operation, a constant pressure 
was maintained in the mains by constantly pumping in excess 
of the demand and allowing the surplus to flow away through 
a relief valve. This involved a constant waste. To further 
moderate fluctuations of pressure four large air vessels (additional 
to those at the pumps) were erected. These were 5 feet in 
diameter and 39 feet high, and were kept charged with air by a 
* CoUadon ' compressor. When it became a question of driving* 


the electric station by turbines driven by water from the high 
pressure system, the need of a storage reservoir became pressing. 
At 4 kilometres from Geneva, a site was found at an elevation of 
390 feet above the lake, and it was decided to construct a 
reservoir, capable of storing the discharge of three groups of 
pumps working through the night. The discharge of the pumps 
is 34,000 c. ft. per hour or 442,000 c. ft. for thirteen hours, 
during which, if there were no means of storage, they would be 
put out of action. 

The reservoir is a covered reservoir capable of containing 
453,000 c. ft. It stores, therefore, 5,573 gross h.p. hours 
of energy. Allowing for the loss at the motors driven by the 
pressure water, the reservoir will furnish about 800 effective 
h.p. for five hours. It serves as a perfect regulator of pressure. 
A float with electric signal and recording apparatus shows 
constantly in the pump-house the condition of the reservoir. 

Hydraulic Pressure Relay or Compensating Pressure Regulator. 
The 16-inch pipe main from the pumping-station to the 
reservoir at Bessinges being 
4 kilometres in length, there 
would be a difference of pres- 
sure in the mains in Geneva 
equivalent to the friction of 
8 kilometres of main, according 
as water was being pumped up p^^ ^3 

to or flowing back from the 
reservoir. This would not have been very serious if all the 
motors driven by the water had been supplied by meter. But 
the larger motors are supplied by gauging, the quantity of water 
used being computed from the area of the orifices of discharge. 
Variations of head would have involved a variation of the 
quantity of power developed at the motors amounting to 20 per 
cent. This would have hindered the development of that 
method of estimating the charge for water.* 

To prevent this variation of head Colonel Turrettini devised a 
centrifugal pump relay, shown diagram mat ically in fig. 43, and in 

* Any difficulty in gauging the quantity of water used by large consumers 
,coqld now be obviated by measuring the water directly by the very ingenious 
and accurate *Venturi' meter of Mr. Herschel. It was used at Chicago to 
measure the supply of water £rom the Waukesha Spring to the Exhibition. 


fig. 44, which comes into action aatomatically and increases the 
pressure whenever the water is returning from the reservoir to the 
town. The centrifugal pump, which forms part of the main, is 
driven by a turbine so regulated for speed that a constant pressure 
is obtained on the town side of the pump. The pump receives at 
the maximum 635 c. ft. of water per minute, and can give to 
the stream passing through it an increase of pressure of 30 feet. 
The turbine works with 380 feet of head, and can exert 120 h.p. 

The sluices of the turbine are governed by an automatic 
pressure regulator. The pressure in the main acts on a piston 
controlled by a spring. According to the position of the piston 
the turbine sluices are opened more or less. The movement of 
the piston actuates the valve of an hydraulic relay, which operates 
the turbine sluices. During the filling of the reservoir the 
centrifugal pump is at rest, the water merelj flowing through it. 
When water flows back from the reservoir, the turbine begins to 
drive the pump so as to increase the pressure in the main. The 
arrangement has worked with perfect success. Fig. 44 showp a 
cross-section and plan of the hydraulic relay and pipes connect- 
ing it to the pumping main. 

The Motors used in Geneva, — The original motors used in 
Geneva were * Schmid * pressure engines, and these are still used 
for small powers. They use a quantity of water which depends 
on the speed only, and not on the work done. Hence they are 
uneconomical with light loads. They are convenient and cheap, 
they can be run at any speed, and they act as meters of the 
quantUy of water used. A counter on the pressure engine, 
recording the number of revolutions, gives the means of ascer- 
taining accurately the quantity of water consumed. At full load 
their efficiency is 80 per cent. 

For all larger motors impulse turbines are used. The maxi- 
mum efficiency of these is 75 per cent., and it is not much less 
with light loads. They occupy little space, and can be perfectly 
governed to constant speed by the ingenious relay governors of 
Messrs. Faesch & IMccard. In Geneva, the question of speed 
regulation was found to be an important one. The industries 
connected with watch-making required motors running at con- 
stant speed. 

TJieEleciric Lighting Station, — In 1887 the Geneva city council 
came to an ai'iangement with a companj^for supplying electricity. 



It was part of the arrangement that the company should use 
pressure water supplied by the town, as motive power in all its 
installations. The pressure water is supplied to the company 
by meter, at a price of 2 centimes per metre cube, with a mini- 
mum of 400,000 c. metres annually. This is equivalent to a 
little more than 31, per effective h.p. per annum. The advantage 
to the town is that their pumping machinery can be run con- 
stantly night and day, the energy which would otherwise be 


Scale of Feet 

io CO 

«o so 

Fig. 44. 

wasted being stored. The electric company, on the other hand^ 
get power at a cheap rate, and their turbines being driven by 
high pressure are convenient and cheap, and run at an ex- 
tremely constant speed. 

Under the arrangement an electric station has been installed 
in the old pumping station, no longer required for its original! 
purpose (see fig. 41). There are three impulse turbines of 200 
h.p. each, and each turbine drives two dynamos directly coupled 



to it by ' Raffard ' couplings (fig. i-j)} There is also a 25 h.p. 
turbine and dynamo for day work. It is the system of reservoir 
storage which makes this hydraulic driving of the dynamos 
possible and economical. Thej' could not be driven so con- 
veniently by thd large low pressure turbines in the river, with 
the very varying head which they have to utilise, nor could 
power be spared to drive them except by utilising the motive 
power of the flow of the river through the night. 

Pig. 46. 

The Installation of the * Compagnie Hydro-EIectrique ' at 
Antwerp. — A very remarkable scheme for the hydraulic distri- 
bution of power is being carried out at Antwerp, and the author 

' Tn the * Raffard' couplings two discs on the shaft are connected by India- 
Tuhber bands, which have a small initial tension when the shafts are not 
driving. The figure shows the position of the bands when the shafts are at 
rest and the position they take when driving. The coefficient of elasticity of 
the rubber may be taken at B = 119 lbs. per sq. in. The limiting stress when 
driving should net exceed 50 lbs. per sq. in. The coupling has the advantage, 
electrically, of being an insulating coupling. It has been nsed extensively, 
even for transmitting large amounts of power, and works very satisfactorily. 
It is most suitable for connecting shafts running at least 250 revolutions per 


is indebted to Messrs. Carels Fibres, of Ghent, who have con- 
structed the steam-pumping machinery, for the following details. 
The plans are based on the investigations of the late Professor 
Fran9ois von Rysselbergh, whose studies and improvements, 
especially in electrical science, are well known. 

Von Rysselbergh recognised the inconveniences of the dis- 
tribution of electricity from a single station over a wide area. 
With low tension continuous current, the cost of the distributing 
mains is enormous ; and with high tension, alternating currents, 
and transformers, he believed the dangers to be' serious. He 
sought, therefore, for means of distributing electricity at moderate 
tension without incurring too great a cost in the network of 
mains. He was led to a system which may be briefly describecl 
as follows : — 

(1) Hydraulic transmission is adopted as the primary means 
of distributing motive power over the district. Water under a 
pressure of 770 lbs. per sq. in. is pumped into the distributing 
mains by steam engines. 

(2) The pressure water is used to actuate motors which 
generate motive power for industrial purposes, or which drive 
dynamos producing electricity in electric sub-stations scattered 
at convenient positions over the district. 

(3) The hydraulic motors are turbines of a new and special 
type, of small dimensions, not costly, and easily managed. It 
is stated that the ' von Rysselbergh ' motor will drive a dynamo, 
.80 that the difference of potential at any point of one of the 
electric circuits never varies more than 2 volts from the mean 
pressure of 110 volts. 

The system, therefore, serves a double purpose. It affords a 
means of distributing motive power in a form very convenient 
for small industries. It also permits the convenient distribution 
of the electric generating stations at as many points in the 
district as are required, and consequently it secures a great 
diminution of cost in the electric distributing mains. Von 
Rysselbergh also contemplated the production of a store of 
electricity at the sub-stations during hours when the demand 
for power or electricity was smallest. 

The company formed to carry out von Rysselbergh's plans 
has erected a principal station in the Rue du Chantier, at 
Antwerp. Here are two compound steam-engines on the Sulzer 



tsystem, constracted by Messrs. Carels Prdres. The cut-off is 
variable for the high pressure cylinder, and regulated by the 
governor. It is fixed for the low pressure cylinder. The high 
pressure cylinders are jacketed with boiler steam, the low pressure 
cylinders with steam from the intermediate reservoir. Tlie 
engines are arranged to work with or without condensation. 

Diameter of high pressure cjlinder 27 ins. 

>f tt "*^ II 11 ..... 40 f. 

Stroke 37J „ 

Bevolutions per minute (normal) 60 „ 

The speed can be regulated by the governor to anything 
between 30 and 75 revolutions per minute. The speed is de- 
termined automatically, according to the demand for pressure 
water. In addition, the accumulators govern the engines, so 
that if at any time there is no demand for water the engines 
are stopped, and start again automatically as soon as the flow 
in the mains recommences. The discharge from the pumps is 
13*7 gallons per revolution at 52^ atmospheres. Two engines 
are at present installed, but there is space for a third. The 
pumps are driven by the prolongation backwards of the piston 
rods. The pumps are on the Riedler system, with differential 
action and valves closed mechanically. The pump plungers are 
8 ins, in diameter. The accumulators are of special cast iron, with 
rams 20 ins. in diameter, and 23 feet stroke. The engines, pumps 
and accessories have been constructed by Messrs, Carels Fr6res, 
the finish being perfect, and all the parts rigorously interchange- 
able. The steam is supplied by five boilers, each having two 
furnaces (internal) with corrugated flues and 16 'Galloway ' tubes. 
The steam pressure is 8 atmospheres. The boilera were con- 
structed by the Soci6t6 Anonyme de Chaudronnerie at Li^ge. 

The distributing mains for the pressure water are steel pipes 
of 11-8, 7'5, 5*85, and 51 5 inches diameter. These pipes are 
all tested to 150 to 200 atmospheres. The flanges, T-pieces, 
and junctions are of steel. The distributing mains will have au 
extension of about 7^ miles, and it is intended that there shall 
be about 12 electric sub-stations. Each sub-station will dis- 
tribute electricity over a radius of 1,600 feet by underground 
conduits, which will have an aggregate length of 10 miles. 




Compressed air has been employed in engineering operations 
for a long period, but it is only recently that its capabilities 
have been adequately recognised. The earliest important ap- 
plication of compressed air was for diving. Diving-bells are 
believed to have been used in the sixteenth century, and 
Smeaton in 1786 and Rennie in 1812 used them in important 
andertakings. Cubitt employed compressed air in sinking the 
piers of Rochester Bridge in 1851, and Brunei used a similar 
ihethod at Saltash in 1854. Compressed air was used in 
driving the Thames Tunnel by Brunei, and Barlow employed it 
in the Thames subway. It has been largely used in tunnelling 
operations since that time. The shafk of the Marie colliery at 
Seraing was sunk by means of compressed air by the Soci6t6 
Cockerill in 1856. 

The use of compressed air in transporting goods was sug 
gested by Medhurst in 1810 and Vallance in 1818. Some 
■early pneumatic railways were built. Similar methods have 
been adopted in recent years for transmitting messages and 
parcels in London and Berlin. 

Papin seems to have considered the transmission of motive 
power by a vacuum method in 1688. Triger transmitted 
motive power by compressed air a distance of 750 feet at the 
mines of Chalonnes in 1845. Soon after compressed air was 
used in several collieries. The greatest impetus to the applica- 
tion of compressed air for transmitting power was due to its 
Adoption for working the boring machinery at the Mont Cenis 
Tunnel. M. Sommeiller, in association with M, Kraft, made 
•extensive experiments on compression at the Cockerill works at 
Seraing. On data so obtained the whole of the machinery for 
■compressing, transmitting, and utilising compressed air at Mont 

II 2 


Cenis was designed and constructed at Seraing. At first a 
kind of hydraulic pneumatic ram was used for compression. In 
1861, this was superseded by water-piston compressors driven 
by turbines. The air was transmitted a maximum distance of 
20,000 feet to work the drills. 

The air pressure used was seven atmospheres (105 lbs. per 
sq. in.). There were at Mont Cenis air mot.or9 worked expan- 
sively, the cylinders of which were heated externally to prevent 
freezing. In the construction of the St. Gothard Tunnel, in 
1872, still more powerful air-compressing machinery was 
employed. The compressors were at first designed to be of 
small size, to run at a higli speed, and to be cooled externally. 
But with a short stroke and quick speed there is not time for 
the heat developed by compression to be abstracted through the 
cylinder wall, and a spray injection suggested by Professor 
Colladon was added. 

In 1877 Mekarski used air compressed to twenty-five or 
thirty atmospheres in conjunction with a small amount of high- 
pressure steam to drive tramway cars, and Mekarski was one of 
the first to use compound compressors. 

In 1877 at Vienna and in 1881 at Paris, M. Popp installed 
a system for working and regulating a great number of 
clocks, by impulses of compressed air conveyed in pipes from 
a central station. A demand arose for a supply of the com- 
pressed air for working small motors, and this proved so 
successful that there has been developed in Paris the most 
important system of power distribution hitherto carried out. 
In Paris, motive power is transmitted to industries of every 
kind over a large area by air compressed at a central station, 
and even sub-stations for electric lighting are driven by 
air motors. It is interesting that in Paris a system of distri- 
buting motive power by vacuum, carried out by M. Boudenoot, 
has been successfully in operation since 1885. The motors are 
worked by atmospheric pressure, and exhaust into pipes in 
which a vacuum is maintained by air pumps at a central 
station. A system of pumping sewage at a number of scattered 
sub-stations by compressed air, supplied from a single com- 
pressing station, has been developed by Mr. Isaac Shone, and is 
in operation at several towns in this country and the United 
States, and at Rangoon. 


Compressed-air transmission is a perfectly general method 
of distributing power for all purposes. Whether in any given 
case it is the most advantageous, the least wasteful of power, or 
the cheapest in working cost, depends on various circumstances. 
M. Hanarte believes that it is and will continue to be the most 
economical method of transmission to considerable distances.^ 
The loss in the air mains is very small. The motors worked 
expansively are efficient. The mains can be carried by any 
path, and differences of elevation between the compressing and 
working points do not sensibly affect the result. In hydraulic 
transmission the water must be collected, stored, and in some 
cases filtered, and having actuated a motor, means must be 
found for removing it. But air is everywhere available, and 
can be discharged anywhere without causing trouble. Com- 
pressed air has peculiar advantages in the case of underground 
transmissions. It has been used to replace manual labour in 
situations where hardly any other motive power could have been 
employed. In driving a tunnel at a mine at Sacramento, for 
instance, the cost was reduced to one-half, and the rate of boring 
was three times as fast when compressed-air machinery replaced 
hand labour. In such cases the advantage is so great, even with 
uneconomical machinery, that the inducement to adopt very 
perfect machinery is absent. Hence much of the air-compressing 
plant at mines has been unnecessarily inefficient and wasteful of 
power. In many cases air-compressing plant has been driven 
by water power, and this also has tended to a neglect of the 
conditions necessary for economical working. Mr. Savage 
argues, with reference to the Temi Steel Works,* that the 
common objection to the use of compressed air on the ground of 
waste of power loses much of its force when the compressors are 
driven by an almost costless supply of natural energy such as 
water power. It is unfortunate for the reputation of the system 
of transmission by compressed air that the rough purposes to 
which it has been applied, the indifference to waste of power in 
mining and tunnelling operations, and the preference for simple 
and cheap machines, have delayed and hindered the improve- 
ment of compressed-air plant. 

* ' Transmission du travail k distance par Tair comprim^ ' ; Congres Inter' 
national de Mica/nique Appliquie, Paris, 1893. 

« « Temi Steel Works,* Savage ; Proe. Inst, Civil En^fieen, vol, xcsiii. 


A good deal was done to improve air compressors by 
Sommeiller, by Dubois and Franpois, and by others, in the large 
plants constructed for Mont Cenis, for the St. Gothard works, and 
for some collieries. In the distribution of power in towns still 
farther consideration has been given to the question of economy 
of working. But here again it has been very unfortunate that, 
in both the great installations in Paris and in Birmingham, there 
were conditions of development very unfavourable to the com* 
plete and fair trial of compressed air as a means of transmission. 
It is reasonably certain that with greater attention to scientific 
principles better results are attainable than have hitherto been 
reached in the use of compressed air. 

For the special purposes to which power distribution id 
applied in London, the high pressure hydraulic system has 
great advantages. Where local conditions permit the construc- 
tion of high level reservoirs, a system, like that in Zurich and 
Geneva, of hydraulic distribution is perfectly successful. Bub 
in more numerous cases compressed air is likely to prove prefer- 
able to hydraulic transmission. It is also the most important 
rival of electrical distribution. There are at present compara- 
tively few cases where electrical distribution of power has been 
carried out, and though enough is known of the capabilities of 
electrical transmission to show that it could be adopted on a 
large scale with complete mechanical success, the cost of dis- 
tribution of power by electrical methods is at present very im- 
perfectly determined. For long distance transmission, and 
where cheap overhead conductors can be adopted, no doubt 
electrical methods have an important field of application. But 
up to the present time, and excluding transmissions for lighting, 
an enormously greater amount of power has been actually dis- 
tributed by compressed air than by electricity.* So far as can 
be judged at present, in the case of distribution of power in 
towns, and especially where work has previously been done by 
steam-engines which can be converted into air motors, in such 
cases compressed air is likely to prove a more convenient and 
cheaper means of power distribution than electricity. 

General Considerations on Compressed Air as a means of 

* It has been stated bj Professor Lnpton that six makers of compressed air 
machinery in England have manufactured compressors working to an aggregate 
of 82,300 h^p. 


Dufiributhig Power in Towns.-^The desiderata in a system of 
power distribution in towns may be shortly enumerated as 
follows : — 

(1) The possibility of indefinitely subdividing the power 
distributed and measuring the supply to each consumer. 

(2) Minimum first cost of distributing maius and minimum 
loss of energy in distribution. 

(3) Simplicity, cheapness, and eflSciency of the motors re- 
quired by consumers of power, and especially that the motors 
should require little attendance and involve little risk. 

(4) Freedom from danger to life or property when accidents 
occur to motors or distributing mains. 

(5) Facility of adaptation to various requirements, additional 
to the supply of motive power. This is important, both from 
the additional revenue obtained, and because the more various 
the industrial applications satisfied, the better are the conditions 
of working at the central station. The fluctuations of demand 
are diminished, and the load-line improved. 

A compressed air system meets these conditions on the 
whole more completely than any other system hitherto carried 
out. It has yet to be seen whether electrical distribution meets 
them equally well. Experience in Paris shows how great the 
facility is for subdividing the power in a compressed air system. 
There are motors ranging from 150 h.p. to less than one-tenth 
of a h.p. (45 foot lbs. per second). The majority of the air 
motors are, in fact, of less than one h.p. These can be started 
and stopped by merely opening or closing the supply valve, and 
the measurement of the air used presents no practical difficulty. 
In Paris and in Birmingham the air is measured by meters, which 
are not costly, and which are accurate enough to give satisfac- 
tion. As to the distributing mains, it may be pointed out that 
in an air system no return main is required, the air being dis- 
charged at the working point without creating any nuisance. 
In a steam distribution a return main is desirable, to avoid heat 
loss, and in an electric distribution a return main is necessary. 
Air mains are less costly than hydraulic mains or steam mains. 
Under what conditions they are less costly than electric mains 
is a question yet to be determined. Probably they are less 
costly than electric mains, except in cases where high electric 
pressure can be used, and overhead conductors. M. Solignac. 


has considered the case of the transmission of 70,000 h.p. from 
Billanconrt to the Place de la Concorde at Paris, a distance of 
4^ miles.^ He comes to the conclusion that air mains woald 
cost 112,P00Z., while electric mains worked at 2,000 volts would 
cost 700,000Z.^ Even if the energy were required at the terminus 
in the form of electricity, he concludes that it would cost 20 
per cent, less to transmit by compressed air and generate 
electricity at the terminus, by dynamos driven by air motors, 
than to generate and transmit the electricity from Billancourt. 
As to loss of energy in the mains, electricity has little ad- 
vantage over compressed air. The pressure loss in the maina 

* SoUgnac, * Transport de V^nergie par Tair comprim^ ' ; Congrei Inter' 
national de MScanique Ajtpliquie. Paris, 1893. 

s Estimates widely different from this will be found, especially in com- 
parisons of the cost of electric and air transmissions made by electrical 
engineers. For instance, there is such a comparison in the report by Messrs. 
Zweifel ani Hoffmann on a project for the electrical distribution of power in 
Mulhausen {Memoires oauronnii par la Soeirte Industrielle de Mulhaute, 
November, 1892). The following is the comparison given : — 

Let it be required to transmit 500 h.p. 2,000 metres. The cost of installa- 
tion will be as follows : — 

Air Tran$mi$twn 


For the compressor 12,000 

Two thousand metres of main, 400 mm. in diameter, 

laying included 120,000 

Air motors and re-heaters 65,000 

Fr. 197,000 

Electric Transmission 


For the dynamo 50,000 

Four thoosand metres of cable, 120 mm. in section . 32,000 

Accessories and erection 8,500 

Electric motors 40,500 

Fr. 140,000 

It will be seen that for transmitting 500 h.p. an air main of 16 inches 
diameter is assumed ; but in Paris it has been shown that 90 cubic feet of air 
compressed to 5 atmospheres (450 cubic feet reckoned at atmospheric pressure) 
will yield one h.p. hour, in a not very perfect converted steam engine motor ; 
that is -025 c. ft. of compressed air, or 01 25 c. ft. of air reckoned at atmo- 
spheric pressure, per second per h.p. But a main 16 inches in diameter, at 
velocities actually used, would transmit 1*4 x 50 = 70*0 c. ft. of compressed 
air per second : it woidd therefore carry 2,800 h.p., instead of the 500 whidi 
Messrs. Zweifel and Hoffmann assume. If their estimate of the cost of the 
air main is divided by five, the relative cost of compressed air and electric 
transmission will look very different. 


of a town distribution ia insignificant. In the Paris system the 
principal mains have an extension of 55,000 metres (34 miles). 
The loss of pressure between St.-Fargeau and the most distant 
point of the main rarely reaches 8 lbs. per sq. in. The safety 
•of an air main is obvious, and even a leakage or burst of the 
main is much less serious and attended with less damage than 
that of a water or steam main. Air leakage is less dangerous 
than electric leakage. When an air distribution is introduced 
in a town, power users do not require new plant and need incur 
no outlay for motors. The boilers, with all their attendant dis- 
advantages of stoking, removal of ashes, cleaning, and risk of 
•explosion, are dispensed with, and the steam-engine, with little 
alteration, serves as an air motor. If an electric system is 
introduced, the old motors must be removed and new motors 
purchased. Further, if electric motors are themselves of high 
efficiency, they run at a high speed, and in most cases there 
is considerable loss in the gearing required to adapt them to 
-ordinary purposes. 

In regard to adaptability to various requirements, compressed 
^r is in a very advantageous position. Electricity supplies 
power and light, but it cannot be used for supplying heat, except 
at a cost prohibitive in most applications. Gas supplies heat, 
power, and light. But for lighting it is open to obvious objec- 
tions, and for heating and power it is expensive. Pressure water 
43upplies power, and indirectly light, if a motor is used to drive a 
dynamo. But except where cheap water power is the original 
«ource of energy, it is too expensive for most purposes where 
motive power is required. Steam supplies heat, motive power, 
and indirectly light if a steam motor is used to drive a dynamo. 
But it is more expensive than compressed air, and involves more 
risk and attention. Compressed air can be supplied so cheaply 
that not only can it be used directly as a source of motive power, 
where that is the commodity required, but it can be advantageously 
used to drive sub-stations and private installations for generating 
-electricity for lighting purposes or for working pumps and 
ventilating-fans. With a water-cushion between the air and 
the lift ram, compressed air is as convenient for working lifts 
as pressure water. It has been used in working cranes at the 
Cockerill works for forty years. Compressed air is not directly 
a source of heat, but used for blowing purposes it is an extremely 


usefal adjunct to famaces. In Birmingham, smiths' fires and 
cupolas have been worked direct from the air mains without 
any blowing machinery. A small jet of high pressure air induces- 
a large stream at lower pressure. In Paris compressed air has 
important applications for refrigerating purposes. Besides large 
refrigerating stores in some restaurants, an air motor is used 
for driving a dynamo for lighting purposes, and the cooled 
exhaust from the air motor is used to cool chambers in which 
food is stored. Lastly, compressed air is already used in work- 
ing tramways, and it appears likely that larger applications of 
this kind are possible. 

General Arrangement of a System of Compressed Air Trans-- 
mission, — The arrangements include— (1) A compressing plant 
driven by steam or water power, with air reservoirs of more or 
less capacity to diminish momentary fluctuations of pressure. 






Fig. 46.— Diagram op Pneumatic System. 

The compressors usually require the addition of cooling arrange- 
ments for absorbing the heat developed in compression. (2) A 
system of air mains for distributing the compiessed air to the 
working points. (3) Air motors driven by the compressed air^ 
and sometimes provided with re-heaters, to increase the work 
.done by the air and diminish the cooling during expansion* 
It is necessary, therefore, to consider the construction of com- 
pressors and their efficiency, the construction of mains and the- 
losses in transmission, and the construction of air motors and 
their efficiency. 

Fig. 46 shows diagrammatically the arrangement of a system^ 
of compressed-air transmission, with one motor system driven 
from the main. The circuit is, of course, an open one, no- 
return main being used. 

Properties of Air, — Let p be the pressure (absolute) in lbs.. 
per sq. ft., v the volume of a pound in c. ft., and T the absolute 


temperature in Fahrenheit degrees. These quantities are con- 
nected by the relation 


Let P,^ v^ Tft be corresponding quantities for air at ordinary 
atmospheric temperature and pressure. It will be assumed that 
ordinary atmospheric air before compression is at a tempera- 
ture of 60° Fahr. or 521** absolute, and at a pressure of 
2,116*3 lbs. per sq. ft. For these conditions 

p, v,=53-2 X 521 = 27,710 ft. lbs., 

and the volume of a pound of air is 

Va = 13-09 c. ft. 

Action in the Compressor. — Consider for simplicity a single- 
acting compressor which receives and discharges a pound of air 
in each revolution, and let the effects of clearance and the 
resistances of the passages be neglected. 

Let P|, Vp T| be the absolute pressure in lbs. per sq. ft., the 
volume of a pound in c. ft., and the absolute temperature of the 
air after compression ; 

Pa> v^j Ta the same quantities for air before compression ; 

p^ and j?a will be used for the corresponding pressures in 
lbs. per sq. inch ; 

r = Va/Vj is the ratio of compression, a quantity determined 
by the mechanical construction of the compressor ; 

p = Pj/p^ = pjp^ may be termed the compression-pressure 
ratio. It depends on r, and also on the thermo-dynamic con- 
ditions of the compression. 

Fig. 47 shows the indicator diagram of such a compressor. 
During the suction stroke a volume v. at pressure Pa is drawn 
into the compressor cylinder. During compression to the volume 
Vp the pressure rises according to some law expressed by the 
curve D G. Finally, the air is expelled into the mains at the 
pressure p,. In general, the compression curve will lie between 
two curves D f, d G, corresponding to two limiting cases. If 
heat is abstracted from the air during compression, so that the 
temperature remains constant, the compression curve will be 
the isothermi^l d f defined by the relation 
p V = constant 

= p. y» = 27,710. 



If no heat is added or abstracted during compression the tem- 
perature of the air will rise, and the compression curve will be 
the adiabatic D G defined by the relation 

p VY = constant, 
where 7 = 1*41. 

In ordinary compressors the curve lies between D f and D G, 
and approximates suflSciently to a curve defined by the relation 

P v° = constant, 
n having a value between 1 and 1*41. 

Fig. 47. 

The whole work of a complete double stroke consists of three 
parts : — (1) The work o a D E of the atmosphere on the piston 
during the suction stroke ; (2) the absolute work of compression 
E D c H ; (3) the work of expulsion of the air into the mains 
H c B o. The efiective work is the sum of these 


that is, the shaded area a B c D. 



Case of Isothermal Compression, — It will be shown presently 
that the most economical compressor mechanically would be one 
in which heat is abstracted during compression, so that the 
compression is isothermal. In that case the effective work is 
(fig. 48), since p v = constant, 

-PaV» + PaV,l0ge^-^-fPlVi 

= p. V, 

* a 

or exactly equal to the absolute work of compression H F D E. 
But the heat abstracted during compression is equal to the same 

quantity. Hence the curious result is arrived at that in the 
most economical compression, the effective work of compression 
is entirely abstracted as heat and wasted. All the compression 
does is to put the air in a condition to do work in a motor at 
the expense of its intrinsic energy. In that way there is 
obtained an amount of work nearly equal to the work done in \ 
compression. But the work in the motor is not strictly the 
restoration of the energy expended in the compressor, but ^ 


'energy borrowed firom the air. Hence the conditions of trans- 
mission of power by compressed air are different from those of 
transmission by pressure water. 

Ome of Adiabaiic Compression. — ^The volame of one pound 
of air at the final pressure P^ will be 

v.=v.(^.)U 1309^7" 
The absolute work of adisbatic compression is per poand of air 

Sence the effective work in one revolution of the compressor 
{abgd, fig. 47) is 

P V — P V 

^1 >1 " '^ -4- p V P V 

_ I ^ * 1 ^1 *^» ^* 


= 95630 [p<^-*« - 1] foot lbs. 

Case of Partially Cooled Compression. — The general equation 
for the work expended in compression is the same as for 
adiabatic compression, if n is substituted for 7.^ If the index 
of the expansion curve is ti = 1-25, the work expended per 
pound of air is 

138550 [>^» - 1] foot lbs. 

Rise of Temperature dariiuj Compression. — For isothermal 
compression the temperature is constant. In any other case 


T, = 521 p u 

For adiabatic compression substitute 7 for n. The rise of tem- 
perature is considerable, as the following table shows : — 

* Let Pa Va Ta correspond to the initial, and P, v^ T, to the final, con- 
^itionB in any compressor. Then 

^ ^ l og (P,/PO 
lojf (Va/V,) 

_log (p,/pO 

log (P,/Pa) + log (T^'T,) 




Ibfl. per iq. in. 


Temperature reached Id compreaaion 



PartiaUj oooled 


11 = 1-41 







• 226 






! 307 


Unnecessary Waste of Work in Heating the Air in the Comr- 
'pressor. — If the compressed air were used in a motor directly 
adjacent to the compressor, in its heated state, there would be no 
necessary loss due to rise of temperature during compression. 
Commonly the air is used at a distance, and has cooled from T, 
to atmospheric temperature t^ and shrunk in volume from B c 
to B F (fig. 47) before reaching the working point. The most 
economical compression for air transmission would be isothermal 
compression. The area F D c represents work expended in the 
compressor which is wasted before the air is used. 

It can be shown that the work wasted in heating the air in 
the compressor above its initial temperature, when the expan- 
sion curve is given by the relation 



p v" = constant, 

n — 1 

-^j[/-u -l]-l0g,p 

i*aVa for air taken from the atmosphere = 27710. The work 
wasted is given in foot pounds per lb. of air compressed. 

Work Wasted in Heating in Compbessob peb Pound of Aib 

Work wasted 


ill lb«. p 


comprcMcU air « 
er 8q. in. 

> Adiabatic 
By gauge compreeaion 
1 a =1-41 






14-7 0077 PaVa 
441 1 0322 „ 
73-6 0-5«2 ,. 

ParUall^ cooled 


n = l-25 

0052 P^Va 
0-209 „ 
0-363 „ 


Tt will be seen that the loss increases rapidly, almost as the 
square of the ratio P|/Pa. This rapid increase of heating loss 
has led many constructors to advise the use of very low working 
pressures in compressed-air transmissions. But that involves 
an oversight. The increased loss at the compressor, due to a 
higher working pressure, is partly balanced by an increased 
eflSciency of the air motor, so that low working pressures are 
not necessarily most economical for the whole system.' 

Efficiency of Compressor and Motor Combined, — If the air is 
compressed isothermally, as it very nearly is in modem compound 
compressors, then the question arises how much work is 
obtained in the compressed air motor. If the air is used non- 
expansively, as in the smaller motors used in Paris, and in much 
compressed-air plant used for rock-drilling and similar purposes, 
the work done in the motor, neglecting valve resistances and 
friction, is simply the admission work Pj Vp less the work of 
expulsion, p^ v^ That is, the work obtained is 


Then the efficiency of combined motor and compressor, with the 
most perfect system of compression, but with non-expansive 
motors, and neglecting all subsidiary losses due to clearance, 
resistance in mains, and mechanical friction of compressors and 
motors, would be 

1 ?2 
V = 

p^Pj=:2 3 4 6 8 10 atmospheres 
ri =072 0-61 0-54 0'46 0*42 039; 

the practically realised efficiency would be, of course, materially 
less than this. 

The efficiency diminishes rapidly with increase of initial 
pressure, and it is the use of bad and inefficient air plants of this 
kind which has given compressed air a bad name, and made 
engineers hesitate to adopt high working pressures. To obtain 
good results the air must be used expansively. 

Suppose the air compressed isothermally and then expanded 



in the motyor down to atmospheric pressure without gain or loss 
of heat, nearly realisable conditions, and let clearance and loss 
in mains be neglected, as well as friction. Then the indicator 
diagram of the motor (fig. 48) will be abfm, the expansion curve 
being an adiabatic fm, and the air cooling during expansion. 
The efficiency of the whole arrangement will be the ratio of the 
area abfm to the area abfd. It is easy to show that that 
efficiency will be given by the equation 


^ = 

\} - (r;)n 

27,710 log. *-' 


Bbsultant Effioibnoy of Compbessob and Motob 

Working: preaeare, Ibe. per sq. in. 


1 By gauge 


i 14-7 
i 441 
i 73-6 
1 102-9 








It is obvious from the diagram that, even if the ratio of expan- 
sion in the motor is not quite great enough to reduce the air to 
atmospheric pressure, the loss of work is not very great. 

Practically, it is necessary to use gauge pressures of 45 lbs. 
per sq. in. at least, that the machinery may not be too cumbrous. 
The calculation shows that when using the air properly much 
higher working pressures may be adopted without serious 
increased loss of efficiency. 

Methods of Cooling the Air during Compression, — Various 
means have been adopted to cool the air during compression, 
and so to reduce the unnecessary heating loss. Sommeiller 
adopted water pistons, the air being displaced by a mass of 
water driven by an ordinary solid piston. The water directly 
absorbs part of the heat, and the cylinder walls are kept cool 
and moist. In the early compressors of this type very good 
results as regards efficiency were obtained. But the Mont 
Cenis compressors were worked at a slow piston speed, and were 



cumbrous and expensive for the amount of work done. It may 
be stated, however, that Mr. Leavitt still uses water piston 
oompressors at the Calumet mines, and by giving suitable form 
to the pistons and passages he succeeds in running them at a 
good speed, and obtains a very satisfactory efficiency. M. 
Hanarte also, in France, has constructed water piston compressors, 
with paraboloidal chambers in which the energy of the water 
pistons is quietly diminished, and shock and loss of energy are 

At the St. Gothard Tunnel works compressors of a less 
effective type were adopted. The compressor cylinder was dry, 
but was surrounded by a water-jacket. Air does not part with 
heat readily to a metal surface, and the heat produced was very 
imperfectly abstracted. Later a method proposed by Professor 
Colladon was adopted. * Water was injected in a fine spray into 
the cylinders, and a much more powerful cooling action was ob- 
tained. To some extent, however, the cooling is deceptive, as 
it takes place late in the stroke and during discharge of the air, 
and the compression curve is not lowered as much as it 
should be. 

In order that the expansion may be isothermal, an amount 

of heat must be removed equal to the work of compression. 

That is— 

P. V. loge — foot lbs. 

= 35-61 loge P Th. U. 

If there is only a small rise of temperature, the heat removed 
will not be very different. Suppose the injection water received 
at 60° and discharged at 100**. Each pound will remove 40 
Th. U. Hence the' amount of injection water in pounds 'per 
pound of air compressed must be 

0-885 loge p' 

, Pounds of injection water 

P Pi/P^ pep pound of air 

2 ' -61 

4 1-23 

6 1-59 

8 1-84 

10 204 



Compound Compressors, — It has been shown that the heating 
of the air during compression is a serious cause of loss of 
efficiency, and that the means adopted to cool the air during 
compression by water jackets or spray injection are not perfectly 
satisfactory. A more efficient means of approximating to 
isothermal compression is to carry out the compression in stages, 
in a compound compressor, and to cool the air to atmospheric 
temperature in intercoolers, between each stage of the com- 
pression. Compound compressors were first used by Mekarski 
and others in cases where high pressures were required. Mr. 

Fig. 49. 

Northcott made a compound compressor with intercooler in 
1878. The Norwalk Company in America constructed com- 
pound compressors for mines in 1880, apparently at first with 
a view of equalising the effort during the stroke. In 1881, 
however, they introduced an intercooler between the high and 
low pressure cylinders. This was made like a surface condenser, 
with thin brass pipes through which cooling water circulated. 

The effect of the intercooler is very important in reducing 
the heating loss. The compound diagram is shown in fig. 49. 
Air is compressed from p^ to p^ in the low pressure cylinder. Then 

N 2 


in passing through the intercooler it shrinks in volume from 
H E to H D, the point D being on the isothermal, if the cooling 
surface is sufficient. It is then further compressed top^ in the 
high pressure cylinder. The worksaved byintermediatecoolingis 
the area E D c G. The adoption of these intercoolers is important, 
because it removes the chief objection to the adoption of high 
working pressures in air transmissions. 

Let Ta be the absolute temperature at which air is admitted 
to either cylinder ; Tj and T, the temperatures at the end of each 
stage of compression ; pa? Pp P^j ^^^ corresponding pressures in 
lbs. per sq. in. Then, if the compression is adiabatic. 

T \i>,-' 

T. ^« nO« 

Tj = T,* ( ^ j = constant. 

The final temperature will be lowest if T, =T,, whence — 

which determines the proportions of the cylinders. The total 
work of compression for the two stages ia per ponnd of air — 



The ratio of the volumes of the cylinders, if the above condition 
is satisfied, is simply 

Pil 2\=V(pJp^)' 

The Influence of Clewrance Space in Compressors, — ^The clear- 
ance space diminishes the amount of air delivered, so that a 
larger compressor is necessary for a given output. The air com- 
pressed in the clearance space to pj expands as the piston 
returns, and prevents the opening of the suction valves till its 
pressure has fallen to p^. There is no direct diminution of 
efficiency, because the work of compressing the clearance air is 
given back in the return stroke ; but indirectly it diminishes 


eflSciency, because the larger the compressor the more work is 
wasted in friction. 

Let V be the volume described by the piston per single 
stroke, cv the clearance volume. The expanding air takes heat 
from the cylinder walls, so that the expansion is nearly iso- 
thermal. Then the clearance air cv Bt p^ expands to CPi/i?«) cy 
=pcv during the first part of the return stroke. The amount 
of air in the cylinder at the end of the suction stroke is v (1 + c), 
so that the amount of air which enters and is delivered per 
stroke, reckoned at atmospheric pressure, is — 

v(l + c) — pcv 

=v— cv (p— 1). 

The ratio of volume delivered to volume described by piston 
or volume efficiency is — 


This diminishes as jPj/jPa increases, and becomes zero at some 
pressure given by the relation 

p=Pi = 1+1 

pa C 

For instance, if the clearance is 10 per cent., the compressor 
will furnish no air if worked at 11 atmospheres, and only half 
the volume described by the piston if worked at 6 atmospheres. 
The importance of reducing clearance is therefore obvious. With 
solid pistons in dry compressors some clearance is necessary ; 
but it may be reduced to 1 to 3 per cent, of the volume de- 
scribed by the piston, by careful design. With water-piston 
compressors the clearance may be reduced to zero, though at 
high pressures an equivalent source of loss would arise from air 
absorbed by the water during the compression stroke, and given 
off during the return stroke. In compound compressors the effect 
of clearance is diminished, because for each cjlinier p^jp^ is less 
than in a simple compressor. 

Loss dvs to Delayed Valve Action, — If the valves close too 
late, there is reflux of air into the cylinder, or from the cylinder 
into the atmosphere, producing a diminished volume eflSciency. 
The loss is similar to that due to clearance. If the valves open 
late, there is an excess pressure at the latter part of the com- 



pression stroke, or a diminished pressure daring the suction 
stroke, either involving a loss of work. With automatic valves 
some loss of this kind is unavoidable. With mechanically moved 
valves there is also often some loss from the ports opening and 
closing too gradually. Professor Eiedler's valves have an advan- 
tage in this respect. Generally, in compressors the suction 
passages have an area of one-quarter to one-sixth that of the 
piston, and the delivery passages of one-tenth to one-twelfth. 

Loss due to Heatirig Air during Adjmission. — In some compres- 
sors the air takes heat from the walls of the admission passages. 
The effect is to diminish the weight of air admitted per stroke, 
so that the volume eflSciency of the compressor is diminished. 

It is desirable that the air taken into the cylinder should be 
taken from a place where the air is as cold as possible. The 
following table gives the diminution pf output for different values 
of the temperature of the air in the cylinder at the end of ad- 
mission : — 

Temperature of air at 
end of Buctiou stroke 





Volume of air delivered 

reckoned at 60° and 

atm. pressure 


Loss of volume delivered 
due to heating of air 
admitted per cent. 


The following table gives the results of some experiments on 
the delivery of a * Dubois- Fran9ois ' compressor, with cylinder 
18 inches diameter, 48 inches stroke: — 

Air delivered at atm. 

Piston speed 

Revs, per min. 

pressure and temperature 

ft, per min. 

in i)er cent, of volume 

described by piston 
















Compressor Valves, — In most compressors simple automatic 
or fluid-moved valves are employed. The objection to them is, 
that they create a resistance to the passage of the air, which 



becomes very serious at high speeds. If they are loaded so as 
to close quickly by springs, they do not open enough to let the 
air pass without lois of pressure. In many compressors 
mechanically-moved valves are employed ; for instance, slide or 
' Corliss ' valves moved by an eccentric, or better by cams. The 
compressor can then be run faster, but there is still usually 

Fig. 60. 

resistance, due to the valves opening and closing too slowly. 
Professor Riedler uses valves which open automatically and are 
closed mechanically. Fig. 50 shows one form of these valves. 
This is a flap valve, so set that it has a slight tendency to open. 
It is closed by a lever worked by a cam. 

The following table, giving dimensions of the *Norwalk' 
compressors, may serve as a guide to ordinary proportions : — 

Table of Dimensions of 'Norwalk* Compbessobs 

Diam. I Diam. 

of low I of high Diam. 


oi iow I oi nigii jjiam. T-nirth "**°" 
pressure pressuic of steam Jr^^Uf^ lutions 

air I air cylinder 
cylinder cylinder ins. 
ins. I ins. ' 

ofjt^ke p.. 


city in 
ft. per 









116! 2 

207 1 2^ 

427 1 3 

960 5 

1,659 j 6 

2,686 7 






















These are adapted for air pressures of 60 to 100 lbs. per sq. in. 

Mr. FearsalVs Hydraulic Air Compressing Enghie, — Mr. H. 
D. Pearsail, apparently incited by a study of the early hydraulic 
ram compressors of Sommeiller, has designed a very interesting 
hydraulic compressing engine, in which air is compressed directly 
by a water-column without cylinders and pistons. 



Pig. 51 shows this engine in section. A is the supply pipe 
by which water, with the energy due to its descent from a 
higher level, flows into the apparatus, c is a large cylinder 
valve which, when open, allows the water to flow out into the 
tail-race, and, when shut, forces the column of water to rise into 
the compression chamber m. The column of water in the supply 
pipe is allowed to acquire velocity by 
outflow into the tail-race. The valve c 
is then mechanically closed, and the de- 
scending column expends its energy in 
compi*essing the air in the chamber M, 
and discharging it into the receiver E. 
The cylinder valve C is actuated by a 
small air motor. The chamber M 
empties the water through the cylinder 
valve C, and fills with air through the 
air valve H, which is controlled by a 
float, K. The adjustment is such that the 
column of water can be made to come 
to rest at the instant when it reaches the 
delivery valve-plate. 

Mr. Pearsall claims that very high 
velocities of flow can be permitted with- 
out danger or loss of efficiency. Some 
experiments made with a small appa- 
ratus gave an efficiency of 80 per cent. 
The engine is simple, and there seems 
no reason why it should not have a high 
efficiency. But, till experiments have 
A ^^^ been made on a larger scale, it is im- 

/ ^^»-^^ possible to say what the delivering 

^ capacity of the machine in a given time 

is. Till that is determined, it is un- 
certain whether it would be more costly, or less costly, than 
ordinary compressors worked by turbines. 

Lossea in Transmission. — ^The frictional resistances in a pipe 
conveying fluid are proportional to the density of the fluid. 
Consequently at equal velocities the frictional resistance of air is 
enormously less than that of water. Conversely, air may be 
transmitted in air mains without serious fall of pressure at ten 


or twenty times the velocity practicable with water in water 
mains. Air at 90 lbs. per sq. in. pressure is about 115 times 
lighter than water, and the frictional resistance at equal velocities 
is less than one per cent, of that of water. In air mains there 
is nothing analogous to the hydraulic shock due to changes of 
velocity, which, as well as the friction, leads to a limitation of 
the velocity of water in mains to 3 feet per second in most cases, 
or to 6 feet per second in some cases. 

In air mains velocities of 30 to 50 feet per second are 
allowed without serious frictional loss. In consequence of this 
high velocity, large amounts of power can be transmitted by air 
at moderate pressures, and in mains of moderate dimensions. 

Most of the hydraulic mains of the London Hydraulic Power 
Company are 6 inches in diameter, the pressure is 750 lbs. per 
sq. in., and the velocity 3 feet per second. That corresponds to 
the transmission of 90 effective h.p. by each main. But air at 
45 lbs. pressure per sq. in., with a velocity of 50 feet per second, 
would transmit 150 effective h.p. in a main of the same size. 
The largest high pressure hydraulic mains are 7^ inches in dia- 
meter, and will transmit 140 h.p. But there is hardly any 
limit to the size of air mains, or the amount of power they will 
carry. The new Paris main from the Quai de la Gare to the 
Place de la Concorde, 20 ins. diameter and 7 kilometres in length, 
with air at 90 lbs. pressure per sq. in., transmits 6,000 h.p. 

In the older Paris mains, which were carried through the 
sewers, and which had an exceptionally large number of bends, 
draining boxes and other sources of resistance, the frictional 
resistance, with a velocity of 25 to 30 feet per second, 
amounted to 2 lbs. per sq. in. per mile of main. It would be 
only in very long distance transmissions that the fall of pressure 
in the mains would be large enough to sensibly affect the 
eflSciency of the system.* 

As to the precise way in which a fall of pressure in the 

mains influences the eflSciency there is a word to be said. If 

air enters an air main at 60 lbs. per sq. in. gauge pressure and 

» M. Solignao says: *Dans le r6seau de la Compagnie Parisienne, qui 
compte 55,000 metres de d^veloppement pour la force motrice, la perte de 
charge entre Saint-Fargeau et le point extreme de la conduit e est 'X peine 
d'an kilogramme ' (7 lbs. per sq. in.). This refers to ordinary working and to a 
main forming a network. The statement above gives the result of a direct 
experiment on a continuous main. 


reaches the other end at 55 lbs. gauge pressure, there being a 
fall of pressure of 5 lbs. due to friction, then it is commonly 
stated that five-sixtieths of the energy of the air is wasted. But 
this is altogether erroneous, the statement being based on a false 
hydraulic analogy. With the fall of pressure in the case of air 
there is an expansion of volume which largely compensates for 
the loss of pressure. The intrinsic energy of the air from which 
the work of the air motor is borrowed remains constant. It is 
only because the air motor works against the pressure of the 
atmosphere, that the available energy of air at 55 lbs. is less 
than that of air at 60 lbs. pressure. 

Suppose that a given amount of work can be done by an air 
motor using a ton of air at 60 lbs. gauge pressure, or 75 lbs. 
absolute. The work expended in driving this motor, by a oom- 
pref sor adjacent to it, would be the work of compressing one 
ton of air to 75 lbs. absolute pressure. Now let the compressor 
be removed to a couple of miles' distance, and the air supplied to 
the motor through a main in which there is a fall of pressure of 
5 lbs. per sq. in. To do the same work as before, all that is 
necessary is that a ton of air should be compressed to 80 lbs. 
absolute pressure. The difference between the work of com- 
pressing a ton of air to 80 and to 75 lbs. absolute pressure is the 
loss of work arising out of the friction of the main, though it is 
not rightly described as energy wasted in friction in the main. 
This amounts with fairly good compressors to about 3 per cent. 

In comparatively short distance transmissions, such as those 
in towns, the loss of pressure in the mains is so insignificant 
that it may be neglected. In long distance transmissions an 
accurate estimate of the frictional loss is necessary. The author 
believes that he has shown, using data derived from careful 
experiments on twenty miles of main in Paris, that long distance 
transmission of power by compressed air is perfectly practicable. 
It is possible, with compressors driven by engines working to 
10,000 i.h.p., to transmit the air in a main of not unusual size 
a distance of twenty miles, and to obtain in motors worked by 
the transmitted compressed air from 4,000 to 5,000 i.h.p., if the 
air is used cold; or 6,000 to 7,000 i.h.p., if the air is re-heated 
before use. 

Air Mains. — The air mains are ordinary wrought-iron tubing 
for small sizes, and steel or cast-iron pipes for large sizes. M. 



Solignac gives the following as the dimensions and cost of the 
air mains in Paris. 






Co»>t per yanl. 



6-72 [ '^^'ro^g^* ^^^^ 

9-12 ) 
2410 (In sewer) ] 
21-90 (In earth) 1^'^*^''°^ 

It will be seen that cast-iron mains have been chiefly used 
in Paris. The new 20-inch main is a steel main. In Birming- 
ham the large mains were of riveted wrought-iron. The 
principal difficulty is the jointing of the lengths of main. For 
cast-iron pipes, and for the Birmingham wrought-iron mains, 
sockets with lead joints have been adopted. But the variations 



v±-/0 Bolts 

Fig. 52. 

of temperature in an air main are liable to be greater than in 
water mains. In Birmingham the variations of temperature 
were excessive. The expansion and contraction makes lead 
joints leaky, and the leakage at Birmingham is said to have 
amounted to 45 per cent, of the whole air supply. At Ports- 
mouth flanged joints are used, and they are said to give no 
trouble, but the distance of transmission is comparatively small. 
The only joint quite suitable for air supply is one adopted in 
Paris, and which is of a type described in Ileuleaux's Con- 
fitructeur as the * Normandy ' joint. The lengths of main have 
perfectly plain spigot ends. A kind of double stuffing-box is 
formed over the two pipe ends, the packing consisting of two 
india-rubber rings. Fig. 52 shows the Paris joint. The india- 
rubber is compressed by two rings drawn together by bolts, and 
is so protected from access of air or water that it appears to be 
nearly imperishable. Perfect freedom of expansion and con- 


traction is secured at each joint in the main. In the case of 
the Paris main (except some parts constructed in the sewers at 
an early date) the leakage is almost negligible. This joint 
appears to have been perfectly satisfactory in Paris, and has 
been adopted at Offenbach and in other German systems of 
compressed air distribution. 


The air motor is essentially a reversed compressor. If it 
works without expansion, it is ineflScient; if it works with 
expansion, the air cools and there is, in some cases, trouble from 
the formation of ice in the valve passages. Rounding the edges 
of the ports so that the ice can be pushed away diminishes the 
trouble. Heating the air before it enters the motor, or during 
expansion, entirely obviates it. 

Cooling during Expansion in the Motor, — In order to show 
how important the cooling action during expansion in the motor 
cylinder is, temperature curves have been drawn in fig. 53 for 
expansion from various pressures down to atmospheric pressure, 
that is for what may be termed complete expansion. The full 
curves show the fall of temperature during adiabatic expansion, 
for which 

pyi-41 — constant 

is the law of expansion. The dotted curves show the corre- 
sponding fall of temperature when some heat is supplied during 
expansion, so that 

pyi-26 _ constant 

is the law of expansion. If P, v, Tj are the initial pressure, 
volume of a pound, and absolute temperature, and Pg T, v^ the 
corresponding quantities at any stage of the expansion, then 

where n has the values 1*41 or 1'25, according as the expansion 
is adiabatic, or with heat supplied as assumed above. 

The lower set of curves show the temperature fall when the 
air is initially at a temperature of 60° Fahr. ; the upper set of 
curves, the temperature fall when the air is reheated before use 
to 200° Fahr. The following table gives the principal results. 
The temperature fall is given when the expansion is complete, so 



that the terminal pressure in the motor is one atmosphere. 
Suppose that on account of ice difficulty, or for any other 
reason, it is desired that the terminal temperature should not be 
below 32** Pahr., then the table gives the limiting terminal 
pressure possible under that condition. 

The curves and the table show strikingly how limited is the 
possible range of expansion, with air not re-heated before use, if 




Fio. 63. 

the difficulty of a very low temperature of exhaust is to be 
avoided. On the other hand, they show how much the possible 
range of expansion is increased by moderate re-heating of the 
air before use, when the exhaust does not fall below freezing 
point. The effect of some heat supply during expansion, 
sufficient merely to alter the index of the expansion curve from 
1*41 to 1*25, is not very great on the terminal temperature. 




Initial Temp. 60** 

Pabtially- Heated Expansion 
Initial Temp. 60** 






Temp, after I 

expansion to I 

ODB atm. I 

Pressure after 

exi)ansion to 


- 35° F. i 
-113 I 
-176 i 




Initial Temp, 200° 

I 80° F. • Complete 




Temp, after 

Pressure after 


expansioQ to 
one atm. 

expansion to 
32° P. 


- 8°F. 



- 66 





InUial Ibmp. 200° 



114° F. 




Types of Motors. — Some special rotary motx>rs are used in 
Paris for very small powers. They are not economical, and are 
rather costly. Most air motors are simply non-condensing 
steam engines, working with air instead of steam. In such 
converted machines there is often not inconsiderable loss from 
leakage, especially piston leakage. In a steam engine the 
condensation on the cylinder wall helps to make the piston 
tight. Air motors are not in the same position, and extra care 
should be taken to prevent leakage. 

Several methods may be adopted to diminish the cooling 
diflSculty and to increase the eflSciency of the motor ; one is to 
inject warm water into the motor cylinder in spray. The air 
takes heat from the water during expansion, reducing it in 
temperature to 32°. The good effect due to injecting even 
moderately warm water is considerable. The following short 
table gives the heat which must be given to the air to keep its 
temperature, during expansion down to atmospheric pressure, 
from falling below 32° ; also the number of pounds of water 
which must be supplied per pound of air, if only the sensible 
heat of the water is utilised : — 

Absolute pressure 

of air, 

lbs. per sq. in. 

Heat requirerl | 
per pound of 
1 airlnTh.U. 


FoanOs of water per pound o 
supplied at 

1-23 -78 

f air, water 







•90 ' 





116 1 


1 160 


1 2-35 

136 i 


r 176 1 


1 2-60 

1-50 1 


Injection of Steam into the Cylinder of the Air Motor. — Steam 
has an enormous advantage over warm water as a means of 
diminisliing cooling during expansion, because it enters in a 
form extremely convenient for distributing the heat throughout 
the mass of air, and because the steam gives up its latent heat 
to the air. If we suppose the steam merely at atmospheric press- 
ure, which is accurate enough for an approximate calculation, 
then each pound of steam received at 212° and rejected at 32° 
will give up 1,146 Th. U. Hence, making the same calculation 
as before, the following table gives the weight of steam which 
must be injected per pound of air to prevent the temperature in 
the motor cylinder from falling below freezing point. 

Absolute pressare of air Heat required per pound 
in lbs. per aq. in. of air in Tli. U. 

Founds of steam required 
per pound of air 


29-4 53 

58-8 I 106 I -093 

88-2 137 -120 

117-6 I 160 , -140 

147-0 176 -154 

Re-heating the Air before Use. — If the air is heated before 
entering the motor, the practical diflSculty due to the cooling in 
expansion is entirely obviated, and an increase of the eflSciency 
of the motor is obtained, which is of the greatest economical 
importance. The air when heated expands, and less air is used 
per stroke. Whether it is economical or the reverse to heat the 
air before use depends on this: whether the additional work 
obtained is more or less valuable than the coal expended. 
Experience shows that it is extremely advantageous eco- 
nomically to re-heat the air. The heat supplied is used with 
great eflSciency, and a larger fraction of it is converted inta 
work than in ordinary heat engines. Further, very small, easily 
managed, and simple re-heating apparatus can be employed. A 
simple coil of pipes, with a small furnace capable of heating the 
air current 300° Fahr., may increase the work done per pound 
of air by 25 or 30 per cent. The heat, according to the experi- 
ments of Riedler and Gutermuth, is used five or six times as 
eflSciently as heat supplied to a good steam engine. 

Fig. 54 shows a simple form of re-heating oven. The com- 
pressed air passes through a double spiral pipe 0. The furnace 



gases rise through the centre of the coil and descend on the out- 
side in a cast-iron casing, with a spiral diaphragm or rib. The 
grate is at F. The air discharged from the motor M may be used 
to create a chimney draught. This has the advantage that the 
draught varies with the amount of work done. As air does not 

readily take up heat from metal surfaces, it is advantageous to 
introduce a small quantity of water into the spiral pipe of the 
re-heater. The water is evaporated into steam, and in the motor 
the steam condenses, giving back its latent heat to the expanding 
air. The water may be supplied from a reservoir above the 


oven, to which the air pressure is admitted so that the water 
descends into the heater by gravity. The reservoir can be re- 
filled by shutting off the air pressure. Steam thus used is 
extremely efficient in increasing the work done by the air, and 
probably the moisture in the cylinder helps to prevent wear and 

In some simple re-heaters tested by Professor Gutermuth in 
Paris, the air was heated from temperatures of 45® to 122° up 
to temperatures of 224° to 363°. From 8,035 to 10,070 Th. U. 
were given to the air per pound of coal used . About 5,200 Th. U. 
were transmitted to the air per hour per square foot of heating 

Combination of a Gas Motor and Air Engine, — In a scheme 
for distributing power, chiefly by compressed air, for the town 
of Dresden, Dr. Pr6ll proposed to work an electric lighting 
station partly by air motors and partly by gas engines. ■ The 
ordinary re-heating apparatus for air motors is not very convenient 
in this case, in consequence of the great variation in the demand 
for power. Hence Dr. Pr5ll adopted the plan of combining 
gas engines with air motors. The gas engine is itself a very 
efficient and convenient motor for an electric lighting station, 
because it can be put in action or stopped according to the 
variation in the demand for power, and there is no waste like 
that due to keeping boilers in steam ready for use. But in 
gas engines a very large fraction of the heat developed is 
necessarily wasted in the water-jacket. Dr. Pr5ll proposed 
to abolish the water-jacket and to take the compressed air 
through the gas-engine jackets to re-heat it on its way to the 
air motors. In addition, the hot gases rejected from the gas 
engine were to be used in the jacket of the air motors. Un- 
doubtedly by the combination of the gas engine and the air 
motor a quite remarkable thermal efficiency could be obtained. 

It may be questioned whether it would not be better to take 
the exhaust of the gas engine directly into the air current. Then 
the gas engine would work with a heavy back pressure, but the 
work BO lost would be recovered in the air motor. In a paper 
on compressed air,^ the author suggested re-heating by the 
burning of gas in the air current, so that the whole of the heat 
would be utilised without chimney losses. Some attempts have 

* * Transmission of Power by Compressed Air/ Froc, Imt, C-E, vol.xciii 



since been made in this direction in America. Fig. 55 shows a 
small petroleum burner used in the air main supplying com- 
pressed air to rock-drilling machinery. 

Meters for Measuring Air supplied to Consumers. — Various 
types of meters have been used in compressed air systems. 
Veiy accurate displacement or positive meters can be constructed, 
but they are costly. Hence inferential meters, which are 
virtually air turbines driven by the air current, are more 
commonly used. Fig. 56 shows an arrangement designed by 
Mr. Abrahams, of the Birmingham Compressed Air Company, 
which is stated to have worked with an accuracy within 1 per 
cent. With a simple fan or turbine driven by the air current 
the velocity of the meter is not proportional to that of the air 
current, in consequence of the friction of the meter. If set 
to be right at a mean velocity, it over- 
registers with a fast current and under- 
registers with a slow current. Mr. 
Abrahams added a kind of pendulum 
governor, the balls being replaced by 
hemispherical cups. The governor 
creates a resistance increasing with the 
radius of the circle in which the cups 
revolve, and therefore with the speed 
of the meter. This extra resistance may be made to balance 
the tendency to over-register. 

Cost of Working with Compressed Air, — Air motors can be 
obtained erected complete for two-thirds of the cost of a steam- 
engine and boiler. In the very imperfect small rotary motors 
in Paris, the K?onsumption of air compressed to five atmospheres 
(75 lbs. per sq. in.) is 750 to 850 c. ft. of air at atmospheric 
pressure per effective h.p. hour. Old steam-engines converted 
to air motors use 450 c. ft. of air at atmospheric pressure per 
effective h.p. hour. 

Now the new compound air compressors at Paris compress 
a cubic metre of air (at atmospheric pressure) to more than 
six atmospheres for 0*4 centime. That is l-08cZ. per 1,000 
c. ft. If the selling price is taken at 2d, per 1,000 c. ft., 
to allow for interest and depreciation on plant, this would 
correspond to 1*66?. per effective h.p. hour for air used in the 
inefficient rotary motors, or -^j^l, per effective h.p. hour for 



air used in the converted steam-engines of tolerably good 
efficiency. In the latter case the cost amounts to 111. bs. per 
eflFective h.p. per year of 3,000 working hours. Allowing for 
interest on the cost of motor and wages, the cost per h.p. per 
annum would be about 131. Better results than this may be 
expected when air motors are constructed as carefully as air 

The prices charged for air in Paris have not been very 

DfAl 9f COUftfEn 

Wheel Vanes 
Guide Vanes 


Fig. 56. 

authoritatively published. It has been stated that 1-5 centime 
per cubic metre is charged. This would make the cost of power 
about double that calculated above. But the Paris prices were 
settled in the early days, when very extravagant and wasteful 
compressing plant was in operation. 

Distribution of Power by Compressed Air at the Works of the 
Societe Gockerill at Seraing. — The great works at Seraing may 
be considered the birthplace of modem compressed-air machinery. 
The compressed-air plant for the Mont Cenis Tunnel works was 
made at the Cockerill works, having been designed and con- 



I stmcted under the direction of Mr. J. Kraft, who is now at the 
head of the engineering staff of the works. In conjunction 
with M. Sommeiller, Mr. Kraft carried out extensive experiments 
on the efficiency of air-compressing machines, in order to obtain 
the necessary data as a guide in attacking what was then a new 
I problem. CompreRsed air in mines was first and is still ex- 
tensively used at the Marihaye collieries at Seraing. Further, 
since 1854 compressed air has been used in the engine works 
of the Soci6t6 Cockerill for working cranes.* With regard to 
this last application, Mr. Kraft states that ' it might be expected 
that the losses of power incurred in the production and the 
utilisation of compressed air would cause it to be rejected as a 
motive power. But in many cases it is not so ; for instance, for 
a series of cranes, machines working only at intervals. Where 
steam is used, enormous losses are caused by condensation in 
the pipes, and expansion and condensation can hardly be used 
in the engines. Whereas, on the other hand, compressed air 
can be produced by high-class engines consuming very little 
coal. In this way the loss incurred by employing air may be 
compensated for. For a set of cranes like those at the Cockerill 
works or at Portsmouth Dockyard, steam cannot compete with 
air. The principal rival of compressed air is water, and there 
are many cases where water is to be preferred. For cranes 
placed in the open air in cold countries the great impediment to 
the use of water is frost.' For the installation of a number of 
cranes in the open air, along a quay wall, Mr. Kraft thinks 
that air is preferable, as in the case of Portsmouth Dockyard. 
The compressed-air machinery in the engine works at Seraing 
in 1885 (and still in use) consisted of the following machines : 

(a) Air-compressing engine with two cylinders. Diam. 
of steam cylinders and air cylinders, 13-78 ins. ; stroke, 29*53 ins. 
Revolutions per minute, 26. 

(h) Air reservoirs, two ; length 36 ft., diameter 6^ ft. Maxi- 
mum pressure, 5 atmospheres. Diameter of pipes, 2 ins. 

(c) 40-ton travelling crane with two double-cylinder air 
motors. Diameter of cylinders, 453 ins. ; stroke, 709 ins. The 
air is supplied to the traveller by a flexible pipe which coils on to 
or off a drum as the traveller approaches to or recedes from one 

* * Notes on Compressed Air and Machinery for Utilising it/ by John 
Kraft; Proe. Inst. Civil Engineers, vol. Ixxxii. 


end of the building. The crane has worked very satisfactorily, 
and was at work this year (1893). 

(d) Three 4-ton swivel cranes. 

(e) Air motor working hydraulic pumps for wheel-press. 
(/) Twelve 12-ton swivel cranes. 

(g) Two 15-ton swivel cranes. 

The compressing engine stops automatically when the 
pressure reaches 75 lbs. per,and begins working again 
when the pressure falls. The air cranes difier in no respect 
from steam cranes so far as their motors are concerned, and all 
can be worked with a pressure of 45 lbs. per sq. in. Also an 
overhead traveller with a reservoir of air at 90 lbs. has been 
erected in the foundry. The air is supplied to the motor through 
a reducing valve at 40 lbs. per sq. in. 

Compressed Air Plant at Poi'tsnvouth Dockyard J — ^In the 
most modern part of the dockyard the lifting and hauling 
appliances are chiefly worked by compressed air. There are 
two sets of compressors, one working to 90, the other to 200, 
i.h.p. The air is compressed at 60 lbs. pressure per sq. in. into 
eight wrought-iron receivers of 18,000 c. ft. total capacity. 
The air mains have a total length of 14,000 feet and vary 
from 3 inches to 12 inches diameter. There are forty 7-ton 
capstans, five 20-ton cranes, machinery for working seven 
caissons, and numerous penstocks, all driven by the compressed 

At Portsmouth there is both an hydraulic and compressed- 
air distribution of power, and Mr. Corner compares their relative 
advantages for such work as that required in dockyards. He 
concludes that the requirements are best and most economi- 
cally met by a compressed-air distribution. This is largely due 
to the advantage of having a considerable store of energy in 
the receivers. There is also less wear in the air plant, and 
it is less easily put out of working order. 

Compressed Air in Mines. — One of the largest mining plants 
worked by compressed air is that at the Chapin Mine, Michigan.* 
About three miles from Iron Mountain, at Quinesec Falls, on a 
head of 52 feet, 1,700 h.p. is obtained by four turbines. Each 

* I. T. Comer: < Jiifting and Haulinfi: Appliances at Portsmouth Dockjard,* 
J^oc. Tnit. Mech. Eng., 1802. 

» See The Iron and Steel Inslitute in America^ p. 378. 


of these drives two * Rand ' compressors. About 2,500,000 c. ft. 
of air are supplied per day at 60 lbs. per sq. in. gauge pressure. 
From the compressor plant a 24f-inch wrought-iron main, J inch 
thick, extends for three miles, an expansion joint being used at 
every 480 feet. The air main is connected to the machinery 
and to 105 povrer drills at the Chapin Mine, and also to some 
neighbouring mines. Most of the machinery is arranged so 
that by closing one valve and opening another a change can be 
eflTected from working by air to working by steam. 

In his address to the Mechanical Section of the British 
Association at Cardiff, Mr. Foster Brown mentioned the use of 
compressed air at a coal-mine in South Wales. There com- 
pressed air has the advantage that it assists ventilation and is 
safe in an explosive atmosphere. There are two tandem com- 
pound steam-engines, working with steam at 150 lbs. per sq. in. 
and developing 1,600 h.p. There is one air cylinder behind 
each low pressure cylinder, 34 inches diameter and 60 inches 

Comp-essed Air Hant at the Steel Woi'ls, Terni^ Central 
Italy} — The locality of Temi was selected for the steel works 
chiefly because of the large water power available there. From 
a point above the Marmore "Waterfall, water is conveyed to 
the works and distributed in pipes to the turbines. The total 
fall from the Velino to the works is 750 feet. The supply canal 
is 7,217 yards in length, of which 2,900 yards are in tunnel. 
The canal carries water suflScient to develop 8,000 effective h.p. 
Part of the hydraulic power is used directly in driving mills, 
hoists, &c. Part is used to compress air, which is then dis- 
tributed to work the great hammer and other machinery. One 
advantage of the compressed-air system is that there are large 
air reservoirs, holding a supply of energy ready for use, when in 
working the heavy machinery there is a sudden and large 

The compressors are Franfois and Dubois compressors. 
There are four sets of compressors coupled to a common shaft 
carrying a fly-wheel. The compressor cylinders are 31^ inches 
diameter and 47^ inches stroke. They are driven by water 
cylinders 14 inches in diameter. The maximum speed is 24 revo- 

' Savage: * Machinery for the new Steel Works at Terni,' Proc. Inst. C. E., 
vol. zciii. 


lutions per minute. The four sets of compressors work to about 
2,000 h.p., delivering 1,8 7 1,000 c. ft. of air, at an effective pressure 
of 75 lbs. per sq. in., in the twenty-four hours. The air compres- 
sors are cooled by spray injection, the water injected per stroke 
being about f^th of the cylinder volume. The temperature of 
the air leaving the compressors is stated to be 120** to 160° Fahr. 

The great 100-ton hammer is worked by a single-acting 
air cylinder 75^ inches diameter and 16^ feet stroke. The 
100-ton and 150-ton cranes are worked each by two double- 
cylinder air engines with cylinders 7*8 inches in diameter and 
12 inches stroke, running at a maximum speed of 200 revo- 
lutions per minute. 

System of Trajismitting Motive Power l/y Vacuum. — An 
interesting plant for distributing motive power by vacuum was 
established in the Eue Beaubourg, in Paris, by MM. Petit and 
Boudenoot.* The general object in view was the distribution of 
power to small industries. From 1874 M. Petit had the idea of 
transmitting power by vacuum. In 1882 an association was 
formed and machinery erected. Conduits were laid communi- 
cating with the houses of consumers, who paid a rental based on 
the number of rotations of their machines ascertained by a 
counter. The users of power were interested in the success of 
the scheme by participation in the profits. The working hours 
are from 7 a.m. till noon, and from 1 r.M. till 8. A steam- 
engine and exhausting pump of 70 to 80 h.p. was first erected, 
the mains extending 300 to 400 yards. Now there are 
three steam-engines, developing altogether 300 h.p., and 
the mains extend 850 yards. There are about 150 small 
motors on the mains. Part of the power is rented to an 
electrical company. This power is supplied by a fourth engine 
of 100 h. p. M. Boudenoot gives the preference to a vacuum 
system because the cost of machinery is less than for a compressed 
air system. The mains are always dry and do not require drain- 
ing-boxes. Lastly, the efficiency of a vacuum system is, in his 
opinion, greater. M. Boudenoot takes the eflBciency of the 
exhausting pump at 0'93, the mechanical efficiency of the 

' * Transmission de la Force motrice par I'Air rar6fi6,' par M. Max de 
Nansouty, OS-nie Civile 1886. • Distribution de la Force motrice k Domicile 
aa moyen de I'Air rar^fi^/ par M. Boudenoot ; Mem. de la SociHe des laffSnieurs 
CiHh, 1885 and 1889. 


vacuum motors at 0*60, and the eflBciency of the expanding air 
in the motor at 0-85. The eflSciency of transmission he takes at 
0*95. The resultant efiBciency is then 

0-93 X 0-60 X 0-85 x 0-95 = 0-4-5. 

This, for small motors, is a good result. The exhausting 
cylinders make 20 to 50 revolutions per minute, and maintain 
a vacuum of 067 to 0*80 atmosphere. These cylinders have 
spray injection. The motors are constructed to supply 360, 540 
and 900 foot lbs. per second. There is a vacuum reservoir, 
50 inches in diameter and 140 inches in height, attached to each 
motor. The vacuum mains are 10,8,6 and 4 inches in diameter ; 
the house-service pipes are of lead. The vacuum system is 
undoubtedly very convenient and efficient for a domestic system 
of this kind. For transmission of large amounts of power to 
greater distances it is not so well adapted, chiefly on account of 
the size of the mains, pumps and motors which would be necessary. 
The Paris S if stem of Distribution of Power by Compressed 
Air, — The Paris power distribution is at present the largest in 
Eurbpe, but it developed out of very small beginnings. About 
1870, MM. Popp and Resch established, first in Vienna and 
then in Paris, a system for regulating clocks by impulses of 
compressed air. At first, in Paris, there was a central station 
in the Rue Argenteuil, with two small compressors delivering 
air into a receiver at 2 to 3 atmospheres pressure. In a second 
receiver air was maintained at a constant pressure of 1| atmo- 
sphere. Two clocks (one in reserve) actuated a distribution 
valve, allowing air to pass into the mains for 20 seconds in each 
minute. By means of small pipes, laid chiefly in the sewers, the 
air impulses were conveyed to the clocks which were to be 
regulated. At the clock a small bellows lifted a rod at each 
impulse and moved the escapement of the clock. The air mains 
were generally f to J inch in diameter, and the service pipes 
into the houses i to f inch in diameter. These pipes were 
extended over many miles, and in 1889 there were in Paris 
8,000 clocks regulated by the pneumatic arrangements. The 
system proved so successful that a new central station was 
erected in the Rue St.-Fargeau, in Belleville. Down to 1887, 
two small Farcot engines and a beam engine sufficed for the 


Gradually there arose a demand to use the compressed air 
for small motors. An extension of the station was then made, 
and a second installation erected in the Rue St.-Fargeau. This 
consisted of six Davey Paxman compound engines, each driving 
two compressing cylinders. The engines developed 2,000 h.p. 
The compressors were made in Switzerland, on the Blanchard 
system. Soon this plant became insufficient. A third instal- 
lation was erected in the Rue St.-Fargeau in 1889.* This 
consisted of five compound engines and compressors, built by the 
Soci6t^ Cockerill at Seraing, the compressors being on the 
Dubois-Franpois system. These engines developed 2,000 h.p. 
Finally, another central station, at the Quai de la Gare, was 



Fio. 57. 

erected, with engines of 8,000 h.p., and room for extension to 
24,000 h.p. The engines are triple vertical engines, with 
Riedler compound compressors. The air is compressed so much 
more cheaply at the new station, that the old station in the Rue 
St.-Fargeau is no longer worked.* 

At the St.-Fargeau station neither the engines nor the com- 
pressors were of the best type. Their eflfect was to compress 

» See the sketch-map of the air-mains in Paris, fig. 57. 

' For numerous details and data relating to the Paris system, see Xeue 
JCrfahrung ueh&r die Kraftvenorgung von Paris durch Druckl%{ftt von A. 
Hiedler, Berlin, 1891; La Distribution de la Force par V Air comprimi dans 
Paris, par Prof. Riedler, Paris, 1891. See also 'Note sur le Transport de 
TEnergie par TAir comprinit^,* par M. Solignac, Cmigrh InterruUional de 
Afeeanique AppliqtUe, Paris, 1893. 


265 c. ft. of air (20-25 lbs.) at atmospheric pressure to 6 atmo- 
spheres per i.h.p. per hour. The cooling in the compressor 
cylinders was ineffective. The air from the six compressors was 
delivered into eight cylindrical receivers of 1,150 eft. capacity 
each. The air was then distributed by cast-iron mains, 11-8 inches 
in diameter ; thes3 had joints with india-rubber rings, forming 
a kind of stuffing-box and permitting expansion at every pipe 
length without leakage. The mains were laid partly under road- 
ways and partly in sewer subways. They were supplied at 
intervals with automatic draining-boxes. It was remarkable 
that the demand for power from this station so rapidly grew up 
to its full capacity. 

The air motors used are generally of a very simple kind. 
For small powers a simple rotary engine is used; for larger 
powers steam-engines are employed, worked with air instead 
of steam. Where the compressed air enters a building it 
generally passes through a screen, which removes solid impuri- 
ties. Then there is a stop valve and a meter for measuring the 
air. Next there is often a reducing valve, by which the pressure 
is reduced to 4^ atmospheres. In most cases there is a re-heater^ 
often a simple double-walled box of cast iron, in which the air 
circulates and is heated by a coke fire. For a 10-h.p. motor 
this re-heater is about 21 inches in diameter and 33 inches high. 
The amount of coke used is not considerable, about \ lb. per h.p. 
per hour. The air is raised in temperature to 300° Fahr. The 
air motors are very convenient. They can be started at any 
moment, they are free from inconvenience from leakage, heat or 
smell, and they require a minimum of attendance. Often the 
exhaust can be used to cool and ventilate the working rooms. 
The air motors are used for various purposes. At some of the 
theatres and restaurants they drive dynamos for electric lighting. 
At some of the newspaper offices there are motors of 50 and 100 
h.p. driving printing-machines. In workshops there are motors 
driving lathes, saws, polishing, grinding, sewing, and other 
machines. At the Bourse de Commerce the compressed air 
drives dynamos for electric lighting, and also is used to produce 
cold in large refrigerating stores. In many of the restaurants 
air is used for cooling purposes. It is also used to work cranes 
and lifts directly, a water -cushion being used between the work- 
ing cylinder and the lift. 



In the first compressors at the Rue St.-Fargeau about 265 
eft. (20-25 lbs.) 
of air werQ com- 
pressed to 6 at- 
mospheres, per 
ih.p. hour. The 
second installa- 
tion had some- 
what better com- 
pressors. These 
compre s sed 
about 300 c. ft. 
(22-92 lbs.) of air 
per i.h.p. hour. 
Later, permis- 
sion was given 
to Professor Riedler to con- 
vert one of the Cockerill 
compressors into a com- 
pound compressor. After 
the change, 370 c. ft. (28-27 
lbs.) of air were compressed 
to 6 atmospheres per i.h.p. 
hour. In the large new 
compressors, at the Qnai de 
of air are compressed per 
i.h.p. hour. The general 
result of the working of the 
new station is that a cubic 
metre of air is compressed 
to 7 atmospheres for U- 1 cen- 
time,* or about four-tenths 
of the cost of compression at 
the older station. 

In the new station, fig, 
58, the steam-en gi lies are 
vertical triple engines*, work- 

» That is, 883 c. ft. or <mJ lbs. 
of air compressed for one peni>y. 


ing compound compressors with two low and one high pressure 
compressing cylinders. Each engine is of 2,000 h.p.* Four 
have been erected, three being regularly worked and ^ one kept 
in reserve. The new station has been placed on the banks of 
the Seine, where coal and condensing water can be obtained 
cheaply. Steam of 180 lbs. per sq. in. pressure is used, and 
the makers guaranteed that the engines would work with 
1-54 lbs. of coal per i.h.p. hour. The new air main, about 
7 kilometres in length, is 20 inches in diameter. 

Compressed Air System at Offenbach^ near FrankforU-ov^Main, 
A compressed-air distribution of the most improved type has 
been constructed at Offenbach, by the firm of Riedinger, of 
Augsburg. The compressing station has a horizontal compound 
steam-engine with crank-shaft and fly-wheel. The air com- 
pressors are compound , of Riedler*s design. The air is compressed 
to 2 atmospheres in the first cylinder, and then to 6 atmospheres 
in the second. The heat is abstracted in an intercooler between 
the high and low pressure cylinders. The steam cylinders are 
22 and 31 inches diameter and 40 inches stroke. The air 
cylinders are 16 inches and 24 inches diameter and 40 inches 
stroke. Each air cylinder is worked direct from the back of the 
corresponding steam cylinder. The engine is stated to work 
with 15^ lbs. of steam per i.h.p. hour, and the efficiency of the 
compressor is 87 per cent. The engine runs at 75 revolutions, 
or 490 feet of piston speed per minute. The air is distributed 
through 23,000 feet of cast-iron mains with india-rubber ring 
joints. The main, when tested with a pressure of 6f atmospheres, 
kept up for 70 hours, showed a leakage of only 1*6 c. ft. per 
mile per hour. 

Comjyressed Air Tramways at Berae,^ — The difficulty of re- 
placing horse traction by mechanical traction on tramways in 
towns is very considerable. Steam-engine traction has obvious 
inconveniences. Wire-rope traction has been remarkably 
successful in certain cases, but it can only be adopted where a 
large traffic is to be carried. Electric traction with overhead 
conductors (the trolley system) has been adopted very extensively 

> In fig. 68 E indicates the position of the water-pnrifying apparatna ; 
FF the boilers; M the 2,000 h.p. compressors; R the air receivers; C a 
channel for water supply ; A A A sewers for removing water. 

* See a paper by Mr. Preller, En^ne&ring^ February 24, 1893. 


in America, but it does not appear to be suitable for European 
towns. Traction by compressed air is possible, and is an 
interesting case of distribution of power from a central station. 

The first system of compressed-air traction which proved 
successful is that of Mekarski, adopted at Nantes, in Paris, and 
at Berne. The Berne tramway has been working since 1890, 
and has gradients of 5 to 6 per cent. The air is compressed ta 
350 to 450 lbs. per sq. in. in reservoirs of about 75 c. ft. 
capacity, carried by each tramcar. Between the reservoir and 
the motor is a small tank of super-heated water at a temperature 
of about 350*' F., supplied from a boiler working at 100 to 150^ 
lbs. per sq. in. The air circulates through the water, and a 
mixture of air and vaporised water passes to the motors, which 
are coupled to the car-axles. The heat furnished makes it 
possible to work with a considerable ratio of expansion. The 
steam yields its latent heat to the air before exhaust. 

At Berne a water-power station has been established on the 
Aare, at a point where there is a supply of 700 c. ft. per second 
on a fall of 6^ feet. There are three 120 h.p. pressure turbines, 
of which two are used for electric lighting, and one for com- 
pressing air. The power of this turbine is rented to the tram- 
way company at 41. per h.p. per annum. There are four com- 
pound compressors running at 80 revolutions per minute, of 
which three are ordinarily suflBcient, the fourth being in reserve. 
Each compressor has two single-acting cylinders. The low- 
pressure cylinder is 11*8 ins. diameter, and compresses the air 
to 75 lbs. per sq. in. The high pressure is 5*3 ins. diameter, 
and compresses the air to 470 lbs. per sq. in. The air is cooled 
by a water-jacket on the high-pressure cylinder, and by spray 
injection in the low-pressure cylinder. Each compressor 
delivers about 350 lbs. of air per hour. The air passes through 
a separator to drain off the water, and then into two reservoirs 
of 44 c. ft. capacity. A wrought-iron main 1 -3 inch in diameter - V^, • ' ' ^ 
conveys the air to the charging station for the cars. At the ^^^ / ^ 
charging station there are six reservoirs of 44 c. ft. capacity 
each. There are also three steam boilers, two with 35 J and one 
with 97 sq. ft. of heating surface, worked at 100 lbs. per sq. in. 
pressure. These serve to charge the heating chambers on the 
cars with heated water. After each journey the cars are re- 
charged with air and heated water by pipes and valves connected 


with the air reservoirs and air main, and the steam boilers. The 
tramcar has two coupled axles with a wheel base of 5*2 feet, and 
wheels 27i ins. diameter. The air motors have cylinders 51 
ins. diameter, and 8*6 ins. stroke. The total weight of the car 
when fall is 9^ tens. The authorised speed is 11 kilometres 
(6*8 miles) per hour. The maximum initial pressure in the 
cylinders is 176 lbs. per sq. in. On easy gradients it is less. 
Each car carries twelve air-storage cylinders of 18 ins. diameter. 
Aggregate capacity, 75 c. ft. Fully charged they contain 177 
lbs. of air at 30 atmospheres. This is suflScient for a journey of 
6 kilometres (nearly 4 miles). The heating chamber has a 
capacity of 35 c. ft. 

The cars are charged first from the station air reservoirs, and 
then from the pumping main, till the pressure reaches the working 
limit. Then the car is disconnected, and the compressor recharges 
the reservoirs. This change of action is effected without stopping 
the compressor. Reloading a car takes about 15 minutes. The 
cars use from 28 to 42 lbs. of air per car mile, or, on the 
average, 35 lbs. The line is nearly two miles in length. There 
is a 10-minute service. The double journey over the line takes 
40 minutes. The working expenses amount to 8'9d. per car 
mile, and the receipts to 10'7d, per car mile, so that there is 
a return of 4^ per cent, on the capital expended. The system 
at Berne has proved convenient and satisfactory. The cars run 
smoothly, and there is no annoyance of smoke or smell. There 
is a somewhat narrow limit, however, to the distance the cars 
will run without recharging. 

Compressed Air Tramway : Hughes ami Lancaster s System, 
This system has not yet been practically adopted, but an 
experimental car has been built at Chester, and it has features 
which are interesting. The car carries air reservoirs, as in the 
Mekarski system, but the air is used at moderate pressure. The 
distinctive feature of the system is that an air main is laid along 
the tram-line, and arrangements are adopted for recharging 
the car, almost automatically, from the air main at convenient 
distances along the line. 

The air is used at 150 lbs. per square inch pressure, and the 
receivers on the tramcar have a capacity of 100 cubic feet. An air 
main of 3 or J? inches in diameter conveys the air to charging 
valves underground at the side of the tram-rail. As the car 


approaches a charging valve, the conductor, by the movement 
of a lever, lowers a plough, which lifts the hinged lid over the 
charging valve. The charging valve has four radiating discharge 
pipes. A valve on the car has two receiving pipes. By a cam 
action, one of the receiving pipes engages with one of the dis- 
charge pipes on the air main as the car moves over the charging 
box. Ab the car moves on the valve opens, then closing again, 
while the radiating discharge pipes rotate through 90°. As the 
receiving pipe disengages, both it and the discharge pipes are 
left in position to repeat the charging operation. The car may 
be stopped for a few seconds in the position for charging, and 
as it moves on again the cover of the charging box falls, leaving 
the road surface unbroken. 

In the experimental oar the air motor was a Rigg engine; 
but on the trials this was in an imperfect condition, and air was 
lost by leakage. The car, starting with a pressure of about 1 50 lbs. 
per square inch, ran distances of 600 to 1,500 feet with a full load 
of passengers, and on gradients varying from level to 1 in 45. 
In a special test it ran with one charge of air a distance of 
1,230 feet up a steep hill, with gradients of 1 in 130 to 1 in 24. 
From the data obtained in tests made by the author, it appeared 
that with a reservoir capacity of 100 cubic feet the car would run, 
with one charge of air at 170 lbs., distances varying from 7,000 
feet on the level to 750 feet on a gradient of 1 in 25. With a 
more satisfactory air motor these distances would be exceeded. 
This system has the very great advantage that there is no limi- 
tation of the distance which the cars will work. It is almost as 
convenient as an electric system in which the supply of energy 
is continuous, and is free from the difficulties and expense in 
electric systems in which the conductors are underground. 

In running down inclines of sufficient steepness, the motor 
on the Hughes and Lancaster car may be reversed, so as to act 
as an air compressor, and it then recharges the air-receivers. 
The surplus work is thus utilised, and the motor acts as a brake. 

The Shone System of Pumping Sewage by Compressed Air, — 
A very interesting system of pumping sewage at many sub- 
stations in a flat town district has been invented by Mr. I. Shone. 
It has been adopted with success at Henley, at Rangoon, at the 
Chicago Exhibition, and at other places. It is strictly a system 
of distributing power for pumping purposes from a central station, 

I U ^" ! V 

•iP TMt X 



and is a good illustration of the convenience of compressed-air 
transmission for certain purposes. 

In order that sewers may be self-cleansing, they must be as 
small as is consistent with discharging the required quantity, 
and mnst be laid at a gradient which secures a sufficient velocity 
of flow. In an ordinary system of gravitating sewers, the lowest 
point available is adopted for the outfall, and the sewers are laid 
with slopes required to give the necessary velocity of flow. But 
to secure this condition the sewers would in many cases descend 
below the outfall level. Then the sewers must be constructed 
at great depths, and pumping at the outfall must be resorted 
to. The cost of constructing sewers at a great depth and the 
expense of pumping are so serious, that the engineer is tempted 
to adopt larger sewers with smaller slope, and to trust to flush- 
ing and other arrangements to remove the deposits which form 
in such conditions. But large flat sewers entail many evils. 

With ordinary methods of pumping it is usually necessary 
to have a single pumping station, to secure economy of superin- 
tendence. The fundamental principle of the Shone system is to 
have a number of pumping sub-stations conveniently distributed, 
each dealing with the sewage of a definite low area. These sub- 
stations are worked automatically, without superintendence, by 
compressed air distributed from a single central station. A town 
which could not be drained as a whole, by sewers with proper 
gradients, is divided into smaller districts, each of which can be 
sewered with self-cleansing sewers draining to one low point in 
the district, at a level generally not more than 15 feet below the 
surface. At each of these low points there is an ejector station, 
worked by compressed air from the central station. The ejectors 
lift the sewage intermittently into a second system of outfall 
sewers, according to the rate at which it collects in the ejectors. 
The outfall sewers ai-e closed pipes without gratings or man-holes, 
and may be either laid with a uniform gradient, or treated as 
pumping mains and laid without reference to gradient, the 
sewage being forced through them by the ejector. For high 
lifts, in cases where the sewage has to be pumped through a 
rising main to be distributed over land, two or more ejectors 
may be used, each dealing with a part of the lift suitable for the 
air pressure employed. 

The ejector (fig. 59) is a simple form of pump for crude 



unstrained sewage. It is a closed cast-iron tank, with inlet pipe 
and inlet valve and outlet pipe and outlet valve. There is also- 
an air valve worked by a float. The sewage flows in by gravita- 
tion till the ejector is full. At that moment a bell-shaped 
float A rises and opens the air valve E. The compressed air enter- 
ing the ejector closes the inlet valve Cand opens the outlet valve 
D, driving the sewage into the outfall sewer. As the sewage 

FlQ. 59. 

falls, it at last leaves unsupported a cup B which closes the air 
valve and allows the ejector to refill. The whole operation of 
ejection takes only about half a minute. Air mains must be 
laid from the central station to the ejector stations. These 
are ordinary cast-iron pipes 2 J to 4 inches in diameter. 

The author tested the efficiency of the ejector system in 
1888. Three ejectors, of 500 gallons capacity each, were 
arranged to pump on lifts of 12 to 21 feet. The air was cora^ 




pressed by a gas engine. The discharge from the ejectors was 
measured in a carefully gauged tank. 

Trial I. — Total lift from bottom of ejector to point of 
discharge, 24*26 feet. Useful lift from centre of ejector inlet 
pipe to point of discharge, 21*26 feet. Speed of working, 20 dis- 
charges per hour from each of the three ejectors. Air pressure, 
11 lbs. per sq. in. Rate of pumping, 331 gallons per minute. 

Brake h.p. of gas engine 
Work due to useful lift 

», „ „ fall into ejector 
Friction of compressor 
Loss in heating air 
Other lofuies 





Trial II. — Total lift from bottom of ejector to point of 
discharge, 12 -3 3 feet. Useful lift from centre of ejector inlet 
pipe to point of discharge, 9*33 feet. Speed of working, 26 dis- 
charges per hour from each of the three ejectors. Rate of 
pumping, 432 gallons per minute. Air pressure, 6 lbs. 
sq. in. 


Brake h.p. of gas engine 
Work due to useful lift 

„ „ „ fall into ejector 
Friction of compressor 
Loss in heating air 
Other losses 




It will be seen that the resultant efficiency of compressor, 
mains, and ejectors is 49 per cent, in both trials, reckoning on 
the total lift. It is 42 per cent, in the first trial and 38 per 
cent, in the second, reckoning on the useful lift, which is exclusive 
of the drop from the inlet pipe into the ejectors. In any system of 
pumping some loss is unavoidable from a drop into the collecting 
tank from which the sewage is pumped, and the loss is proportion- 
ately greatiCr the less the lift. This is a very satisfactory 
efficiency on such low lifts. The working of the injectors was 
entirely automatic, and they required no attention. 




Loss of Pressure in the Air Mains, — Hitherto information as 
to the resistance of air mains has been scanty. The best experi- 
ments were those made by Mr. Stockalper on the pipes of the 
boring machines at the St. Gothard Tunnel. The new experi- 
ments carried out by Professor Riedler and Professor Gutermuth,^ 
on the air mains in Paris, are therefore of great value. The 
Paris mains are larger than any hitherto tried, and by coupling 
up different mains at night, a length of 10 J miles could be 
experimented on. 

The main for the older Paris compressing station consists of 
cast-iron pipes 11| inches or 98 foot in diameter. It was laid 
partly in the sewers, which involved the use of a good many 
bends. Part of the main in the Rue de Belleville is known to 
be leaky, and there are numerous draining-boxes, siphons, and 
stop-valves which cause resistance. In a more perfectly 
arranged main no doubt the loss of pressure would probably be 
somewhat less than in this old one in Paris. In the following 
investigation Professor Riedler's results are used, but the 
reductions from them and the conclusions deduced are the 
Author's. The formula for the flow of air adopted is that given 
in the Author's paper on * The Motion of Light Carriers in 
Pneumatic Tubes.' ^ 

* This chapter is mainly a reprint, by permission of the Coancil of the 
Institution of Civil Engineers, of a paper by the Aathor in the MinuUi cf 
Proeeedingi, vol. cv. 

' Neue Wrfahrungen iiber die Kraftversorgtmg wm Paris dureh Drucklu/t, 
Berlin, 1891. 

* Minutes qf Proceedings last. C. E., vol. zliii., 1876. Also article on 
* Hydromeobanios,' B/u>gelojfadia Britannieaf ninth edition. Hydraulics, § 84. 

p 2 


Formula for the Flow of Air in Long Pipes, — Let d be the 
diameter and I the length of a pipe in feet ; v the velocity of 
the flaid in feet per second, and h the head lost measured in feet 
of fluid under the given conditions. Then, if the fluid is incom- 

ir^ V Z^ ^^"* ' '^ 

This formula may be used for the flow of air in pipes, and 
indeed has been so used by Mr. Stockalper and others, when the 
variations of pressure and density are small, so that mean values 
can be taken and the variations neglected. Professor Riedler 
uses this equation also, but with the artifice of dividing a long 
main into portions calculated separately — a method very 
cumbrous and not very accurate. 

When air flows along a pipe there is necessarily a fall of 
pressure due to the resistance of the pipe, and consequently the 
volume and velocity of the air increase going along the pipe in 
the direction of motion. The efiect of the resistance is to create 
eddying motions, which, as they subside, give back to the air 
the heat equivalent of the work expended in producing them. 
The result is that, apart from conduction to external bodies, the 
flow is isothermal. Generally, in compressed air systems, the 
air is delivered into the mains at a temperature above that of 
the surrounding earth. The excess of heat is parted with by 
conduction, and the temperature falls to that of the ground, but 
no lower, for there can then be no further loss of heat by con- 

Let p s= the absolute pressure of the air in lbs. per square 

T 5= the absolute temperature. 

G = the weight of the air per cubic foot in lbs. 

V = the volume of the air per lb. in cubic feet. 

Then, f zi x 

' PV=-=CT . . . . (1) 


where c=53-18 for air. Taking the temperature of 60° Fahren- 
heit, so that T=461 + 60=521— 

ct=5318t=27,710. ... (2) 



If air 18 flowing steadily in a pipe, the same weight of air 
flows across every transverse section per second. Hence, if w 
is the weight, in lbs., of air flowing per second, fl the area of a 
cross-section, at which the velocity is u — 

w=Gftw = constant ... (3) 

combining (1) and (8) 

nwP=(jTW .... (3a) 

Fig. 60 represents a short length dl of an air main, between 
transverse sections A^^, a,. Let d be the diameter, fl the cross- 
section, m the hydraulic mean 
radius of the pipe. Let P and u 
be the pressure and velocity at A^, 
p + cZp and u + du the same qpan- 
tities at Aj. Let w be the weight 
of air flowing through the pipe per 
second. The units are feet and 

If in a short time dt the mass 
Aj^, comes to a'^a',, then AqA.\=zu dt and A^A\si(u + du) dt. 

By analogy with liquids the head lost in friction measured 
in feet of fluid is — 


dl ^ 


»» t *\»PP9»n*»9» 9 m9t»9»fWftPPUHt 9m M ^ 

A. a: 

Fig. 60. 


then the head lost is 

and since w dt is the flow through the space considered in the 
time dty the work expended in friction is — 


The change of kinetic energy in the time dt is the difference 
of the kinetic energy of a^a'j and a^a^', that is — 

^'^^^ {(u^duy^u^] 

= —ududtssw dRdt, 


The work of expansion of £ludt cubic feet of air to fl (m -h du) 
dt^ at a pressure initially F, is fip dudt But from (3a) 


u= - — 

du CTW 

and the work done by expansion is — 

— d? dt. 


The work done by gravity is zero if the pipe is horizontal, 
and in most cases may be neglected without great error. The 
work of the pressures on the sections a^Aj is 

2£ludt — (P + ^P) ft (w + du) dt 
= ^(?du + udp)Qdt. 

But if the temperature is constant — 

Ptt= constant 
PcJiA + i/dp=0 

and the work of the pressures is zero. 

Adding together the quantities of work and equating them 
to the change of kinetic energy — 

w dn di^^^^dFdt ^t-wdl dt 
P m 

dB + —dF + t-dl=^0 

p ^m 

— +-^-dP + f'^'=0 . . . (4) 
H HP m ^ ' 

But _CTW 

^"" ftp 



H CTW^ m 


For pipes of uniform section fl and m are constant, for steady- 
motion w is constant, and for isothermal flow T is constant. 
Integrating — 

log H + ^^^- -f f -^ = constant . . (5) 

For Z=0, let h = Hi and pssPj 
„ Z=L, let H=H^ and p=Pj 

logI^+ ^?' (P,^-V)+?-=0 . . (6) 

where Pj is the greater and Pj the less pressure. L is the length 
of transmission. By replacing Hp H, and w — 

Hence the initial velocity in the pipe is- 



and when L is great, log-' is comparatively small compared with 
the other term in the bracket. Then — 

For pipes of circular section m= -, where d is the diameter 

in feet. Let cT=27,710, and let |9,, p^, be the pressures in lbs. 
per square inch. Then — 

«. = ^|222,900^^ P^'-i'^'] . . (7a) 

This equation is easily used. In some cases the approximate 
equation — 

2«i=^M319-07264^i)A/r222,900,'^) . (8) 
may be more convenient. 


If the terminal pressare p^ is required in terms of the initial 
pressure pi, then — 

If from a series of experiments f is to be found — 

f = 222,900 i- ^'»'-?2' 



Variation of Pressure and Velocitif in a Lowj Main, — In order 
to have some idea of the law of variation of pressure and velocity 
in long mains, two cases have been calculated. Taking a main 
of 12 inches diameter, and assuming f= 0*003, the following 
results are obtained 

At Distancbs 

[N Miles peom Obioin op 





8 4 



8 9 


Cask I.— 




Pressure (^abso- 




1070 104-3 


98-5 96-r 92-3 891 


lute) in lbs. 

per sq. in. 

Velocity in main 







29-2' 29-9 31-2; 323 


in ft. per sec. 


Case II.— 

Pressure (abso- 




78-6 61-8 


lute) in lbs. 

per sq. in. 

Velocity in main 

50' 561 


73-2 931 

149-4' 00 ; 1 

in ft. per sec. 

1 1 ' 

1 1 
1 i 

It will be seen that with an initial velocity of 25 feet per 
eecond the pressures decrease comparatively slowly and the 
velocities increase also somewhat slowly. Even at 10 miles the 
pressure is still considerable and the velocity moderate. With 
an initial velocity of 50 feet per second, the variation of pressure 
and velocity is much greater in long mains. Beyond 5 miles 
the pressure becomes too small to be practically available aud 
the velocity enormous. 

These results are plotted in curves in fig. 61. 

Before proceeeding to calculate the value of ^ from the 
experiments of Professor Biedler and Professor Gutermuth on 
the Paris mains, it is necessary to examine the corrections to 
be made for leakage and the special resistances of the draining 



Loss of Air hy Leakage from the Main, — Special experiments 
were made by Professor Riedler and Professor Gutermuth to 

Ar£r LSS. 
PER P£ff 

SEC som 




























1 \ 



t - 



* i 

f J 

' J 

Fig. 61. 

determine the leakage loss in portions of the old Paris main. 
These gave the following results : — 










Station to 




Works to 

PL de la 

Station to 

PI. de la 






Length in miles . 







Initial pressure in 








Terminal pressure 







Loss of pressure 







per honr in at- 


Loss of air in cnbic 







feet per hour, 

reckoned at at- 

mosphere pres- 


Loss of air in cubic 







feet per mile 

per hour 



It is stated that during trials I, II, and IV, considerable 
leakage was known to exist at the Central Station and the 
Rue de Belleville. In all the trials the air consumed by the 
clock system and some small motors not stopped is included. 
Reduced, as shown in the last line, the resulta are consistent, 
if it is assumed that the leakage was really greater in the 
sections included in I, II, and IV than in the other sections. 

Now, with an initial gauge-pressure of 6 atmospheres and an 
initial velocity of 30 feet per second, the main would deliver 
about 600,000 cubic feet of air per hour, reckoned at atmospheric 
pressure. Then the percentage of loss by leakage per mile per 
hour would be as follows : — 

; I 1 II in 

1 1 



Per cent, of air lost i 1-00 106 i 0-38 
by leakage per 
mile per hour | 

0-91 0-40 

Professor Riedler believes that in newer and better laid 
mains the leakage is considerably reduced. In any case, it 
appears that the loss is small and practically negligible except 
in very long transmissions, at least when the main is delivering 
a full supply. It should be remembered, however, that the 
leakage is proportionately greatest in those hours when the 
pressure has to be maintained in the mains, but the demand for 
power is small. Where the demand for power is very variable 
this loss might become an appreciable factor. Looking to the 
uncertainty as to the amount and law of variation of leakage in 
different cases, it will be neglected in this investigation, takings 
the experiments above referred to as a proof that it is not a 
serious quantity even in the Paris installation, and that in newer 
mains laid with the experience now gained it would be still less- 

Corredimi for Loss of Pressure at Draining Tanks, — On some 
of the mains experimented on by Professor Riedler there were 
draining tanks at which there was a sensible loss of pressure* 
The resistance at such a receiver must be analogous to the loss 
at a sudden enlargement of a pipe, and we may therefore write. 



if 1) is the pressure lost, p the absolute pressure of the air, and 
V its velocity — 

where t^ is the coefficient of resistance to be determined by 
experiment. Some experiments were made by Professor Riedler 
on these draining tanks. The following table gives the results 
and the calculated value of ^y. 

Absolute pressure p. 

Velocity of air 

lbs. per sq. In. 












Observed loss of 
pressure at one 
draining tank, 1) 


value of ^d 




These figures are fairly accordant, 
draining tank may be taken as — 

1) = 0-00003 jw« 

and the loss at each 


Experiments by Professor Riedler and Professor Guterm/uth on 
the Resistance of the Paris Air Mains. — ^The old air mains of 
Paris consist of cast-iron pipes 30 centimetres = llf inches = 
0*98 foot in diameter. It was on these that the experiments 
were made. For the new station wrought-iron pipes of 20 
inches diameter are being used. The importance of these 
experiments lies chiefly in the large scale on which they were 
carried out. The volumetric efliciency of the compressors was 
first determined. Then, at different points on the pipe-line, the 
pressure was observed when the compressors were running at 
known speeds. On Sundays the whole pipe-line could be 
coupled up, all work in the city being stopped. The air then 
passed through the southern main and back by the northern 
main, a distance of 10 miles. The main is somewhat complicated, 
having been constructed piecemeal during the gradual extension 
of the enterprise. In the 10-mile length there are four draining 
tanks, twenty-three draining traps or siphons, and forty-two 
stop-valves. The resistance of these is included in the observed 


results. It appears, however, that the most serious additional 
resistance was that of the draining tanks, and as special experi- 
ments were made on these, their resistance can be estimated with 
very approximate accuracy and allowed for. 

The table opposite gives the results of the experiments, re- 
duced to English measures. The columns marked with an asterisk 
are those taken directly from Professor Riedler*s tables. The 
others are deduced from hi^ figures. 

The last column gives the values of f calculated by equa- 
tion (10). 

It will be seen that the values of f are fairly consistent con- 
sidering the difficulties of the observations. The discrepancies 
are irregular, and obviously due to inconsistencies in the data 
of the experiments. For instance, experiments XIII and XIV 
were made on the same main at nearly the same initial 
pressure. In XIV the velocity was sensibly greater than in 
XIII, but the loss of pressure is much greater in XIII than 
in XIV. 

Professor Riedler appears to believe that his experiments on 
the Paris mains show that the friction of air in pipes is consider- 
ably less than it was believed to be, from the results of earlier 
■experiments on a smaller scale. The Author does not think that 
this is so, but that, on the contrary, when properly reduced they 
are consistent with previous results, and that all the known 
experiments fairly support each other. In a paper on the 
* CoeflScient of Friction of Air Flowing in Long Pipes,' * it was 
«hown that Mr. Stockalper's experiments gave the following 

Mean for 0-492-foot pipe 0-00449 

„ 0-666-foot pipe 0-00377 

These experiments, as well as some others given in that 
paper, show that the value of f for air, as is the case for water, 
decreases as the size of the pipe is larger. Though experiments 
on the flow of air are not numerous enough to furnish any very 
trustworthy law, the author gave in 1880 the following expres- 
43ion for the value of f : — 

r=00027(l + 4) 

* Minutes of Proceedingt Inst. C. £., vol. Iziil. p. 29 . 









'|8S8i8888§8 '88 

|oooooooo666 66 

«ppoo«ioaioooooo;ob- CC-liO 

^ * = fc -*^ 

00 I'- «: o » C'l X h' « lo c o o !M 

P ip Ol fH ip ^ -44 

o »b a» 6» M4 do cb 

« eo O t* ;o oa » 

»o ©« w 1-t fH c<» CO 



»0 »o -2« ?o 00 

•?< ^ O «o o> 

o t- 4n 

*i C^ .M 

s § :^ 




OOOOOO — ^^o^o 


S-3 5« aJ5 T3 





00 f^i ^ 

do 6) to 


J3 . ^ 

9J ft^-^ ff ^-' 

04 00 CO p Op b- p 
h- do o b- op -^ «b 

t- b- O O O ^ »- 

2 8 

"N CO 

S 2 






►-» 2 »-H 




o "t O 


* I I I « 

• § § S I 

J 2 5 .2 

C3 3 c t^ 

o .a o iS 

P^ O PK4 cc 









This gives the following values for f : 

Pipe 0*492 feet diameter 
„ 0-656 „ >• 
„ 0-980 „ 


which are only a little different from those deduced from Stock- 
alper's and Riedler's experiments. 

Rounding off slightly the mean of Professor Riedler's results, 
it will be assumed in the remainder of this chapter that, for air 
flowing in pipes of not less than 1 foot diameter, f = 0003. 
Putting this value in equations (7a) and (9) — 

74,300,000 - 




ih-lh\/ |i 74,300;000 




Ijoss of Pressure 2>er Mile of Pipe. — In order to indicate what 
kind of results this formula leads to, the following cases have 
been calculated. 

1. Given the diameter of the pipe, and the initial pressure 
and velocity of the air entering the main, it is required to find 
the loss of pressure in one mile of transmission. By simple trans- 
position, putting L = 5,280 feet — 




Assuming initial velocities of 25, 50, and 100 feet per 
second, and initial pressures of 50, 100, and 200 lbs. absolute, 
the following values are obtained : — 


of pipe, 


Initial velocity 

^i 1 
ft. per sec. i 

Values of terminal pressure j» 

„ when 

linitial pressure 
llost in one mile 






100 1 

25 ! 
50 1 







1 2-4 

1 9-4 


' 4-6 

The percentage of pressure lost in a mile is the same what- 
ever the initial pressure. It must not, however, be assumed that 



the loss in two miles is doable the loss in one mile. The velocity 
increases and the density diminishes going along the main. 

It is clear that when the velocity initially in the main 
exceeds 50 feet per second, the loss of pressure becomes serioas 
even in a distance of a mile. How far that involves a loss ot 
efficiency will be considered later, 

2. Another mode of looking at the question is interesting. 
Suppose the percentage loss of pressure is assumed and the 
initial velocity calculated. In the following table this has been 
done for transmissions to distances of 1, 5, 10, and 20 miles. 

Length of 

Diameter of 

main in 

main in 

in trauunisaion is 



1 j 2i 6 1 10 1 20 




16-7 26-3 370 i 52-6 731 

23-7 37-3 ; 62-4 7M8 , 100-6 


r 1 


7-6 1 11-8 16-5 1 23-5 1 31'8 

10-6 1 16-6 

23-3 , 33-2 44-9 



6-3 8-3 

11-3 1 16-6 22-5 

7-4 11-7 

16-9 1 23-4 31-8 




.8-7 5-9 

8-3 ' 11-8 I 16-0 

6-4 8-3 

11-7 1 16-6 22-6 

Transmission is possible to the longest distances here assumed 
at velocities not impracticably low, with losses of pressure which 
would not hinder efficient use of the air. 

Action of the Air Compressors. — When an air compressor is 
driven by a steam engine, there is a difference between the work 
done by the steam and the work done on the air, due to the 
work expended in friction of the mechanism, and measured by 
the difference of area of the indicator diagrams of the steam 
cylinder and compression cylinder. If the compressor is driven 
by water power there will also be a corresponding loss in friction 
of the mechanism, probably not widely different in amount. It 
will be sufficient here to consider a compressor driven by steam. 

Let u be the work expended, measured on the steam-cylinder 
indicator diagram, and u, the corresponding work shown on the 
compressor-cylinder diagram. Then, if rj^ is the efficiency ot 
the mechanism, 

Ui = i7jU. 

It is convenient at present to take u and v^ to be the work in 
foot-pounds per pound of air compressed. In Professor 


Kennedy's tests of some of the older compressors at Paris, it 
appeared that i;j = 0-81j5. Experiments by Professor Gutermuth 
on the new Eiedler compressor gave -?7i =0*87, a result not widely 

In all compressors there are some losses of work due to 
clearance, to imperfect action of the valves, to leakage, and to 
other causes. But unless the machine is badly constructed 
these need not be large. If the air is compressed to a pressure 
above that in the mains — as happens in some cases — there 
is a loss, for the unbalanced expansion uselessly heats the air ; 
but this again can be kept within narrow limits if the valves act 
promptly and properly. The chief loss of work in compressors 
is due to useless heating of the air. It is not, of course, 
impossible that some of the heat thus generated should be use- 
fully employed. But to make use of it would involve complica- 
tions. It is said that in Birmingham, where the distance of 
transmission was not great, some of the heat reached the motors, 
and then there was an economy in the amount of air used. But 
practically, in most cases, heat given to the air in the compressor 
is in fact wasted before the air is used for motive purposes. A 
perfect compressor for power distribution would be one in which 
the air, taken in at atmospheric pressure p^, should be compressed 
isothennally to an absolute pressure p^ equal to that in the 
mains at the compressing station, and then delivered into the 
mains without valve-resistance. 

Such a machine, working without friction, or clearance, or 
valve-losses, would require for each pound of air compressed an 
amount of work given by the equation — 

Ug = P^\\ log, ^^ foot-lbs., 

where P^ is the atmospheric pressure in lbs. per square foot, and 
Vh the volume of 1 lb. of air at that pressure. 

If, as hitherto assumed, the initial temperature of the air is 
60° Fahrenheit, 

u,= 27,710 log/i. 

Fig. 62 shows the indicator diagram of such an engine. 
CADE is the work done by the atmosphere on the piston in 
the suction stroke, gbcde is the work done on the air in 



the compressing stroke. The work expended in compression is 
therefore the shaded area, a b c d. 

If the air is compressed at constant temperature, the com* 
pression curve is the isothermal d f. The work of compression 
is then the area A D F B. If the air is compressed without any 
cooling during compression, the compression curve is an 
adiabatic such as d g. The work expended in compression is the 

Fio. 62.— Action of Compressor. 

area a d g b. Consequently, in that case the area F g d repre- 
sents the work expended in useless heating of the air. This 
is easily calculated, for it can be shown that the work expended 
per pound of air in adiabatic compression is : — 

,4i -. [i^rM 

Putting 7=1-41 and P„v«=27,710 as before, 




If i;,=Uj/u',, then 17, may be termed the efficiency of the 
compressing process. For simple adiabatic compression, neglect- 
ing clearance and friction losses — 

^j = 


log. Py 

V 0-29 

This may be called the theoretical value of rj^. 


- \ i ' 





14-7 1 3 

— 5 

— 7 

ralue of i|, 


It is clear, therefore, that if the air is not cooled during com- 
pression, the efficiency of the process decreases as the pressure 
to which the air is compressed is greater. Now in most cases 
the cooling arrangements are very imperfect, and the compres- 
sion is nearly adiabatic. Consequently there has been reluctance 
to use high initial pressures, and this diminishes the facility for 

VoiaM£ or Air admitted 

Fig. 63.— Colladox Comprbssob. Revolutions, 104; Work 
wasted, 38-15 pee cent. of useful wobk. 

distributing power by compressed air. As the other losses 
besides that due to heating are serious, the whole efficiency of 
the compressor is considerably less than the values calculated 

Fig. 63 shows an actual diagram from one of the best of the 



older Colladon compressors. The shaded area, bcfa, is the 
useful work of compression, Uj, corresponding to the area 
A BCD in fig. 62. The thick line is the actual indicator 
diagram. It will be seen that the actual diagram is 38*15 per 
cent, larger than the isothermal diagram for the volume of air 
admitted, the difference being due partly to the waste of work in 
heating the air, and partly to valve resistance, causing a loss of 
pressure in the suction stroke and an excess pressure during 
delivery into the mains. In this case 17^ measured from the 
diagram is 072. In several compressors tried by Professor 
Biedler the loss of work in the compressing process was twice 

r VoLuiii "or 'Am AOMirfio 

Fig. 64.— Cockbbill Compressor. 

as great as in this case. Fig. 64 shows a similar diagram for 
one of the Cockerill compressors at Paris. The excess work 
here is 40*2 per cent, of the useful work, and consequently 

Obviously, part of this loss can be saved if the air is cooled 
during compression, and all compressors are provided with some 
means of cooling. Very generally a water jacket to the com- 
pressing cylinder is used, but the action of this is very imperfect. 
At Birmingham, although the compressors were well water- 
jacketed, and though the pressure was only 45 lbs. (gauge), the 
temperature of the air delivered was about 280'' F. The area of 




surface of the cylinder is small compared with the volume of air 
compressed, and air parts with heat to a metal surface slowly. 
Spray injection into the cylinder answers much better in keep- 
ing down the temperature. But it is believed that the result is 
partly deceptive, the cooling going on after the air is completely 
compressed, so that the compression curve is hardly so much 
flattened as might be expected from the temperature at which 
the air is delivered. 

Fig. 65 shows a diagram from a compressor with water 


Fig. 65. 

pistons. Here the cooling is much more effective. The piston 
speed is 180 feet per minute. The excess work is only 19*6 per 
cent, of the useful work, and 1/2=84. 

The only remaining means of reducing the loss of work by 
heating is to compress in two or more stages, and cool the air 
thoroughly between the stages. This intermediate cooling can 
be easily effected, the air being taken through tubular vessels 
presenting any amount of cooling surface that may be required. 

In fig. 66 A B K F is the diagram of isothermal compression, 
as before, f E G is an adiabatic, so that in single-stage com- 



pression, without cooling, KDFG represents the work lost in 
useless heating. Now let it be supposed that the compression 
is effected in two stages. In the first stage, no cooling being 
assumed, the work tKt will be the area hefa. But the air 
is then cooled to its initial temperature, and the volume shrinks 
from H £ to H D. D is a point on the isothermal F K. During 
the second stage DC is the adiabatic compression curve, and 
D K c the work wasted in heating the air. It can now be seen 
that the effect of the intermediate cooling is to reduce the work 
expended in compression by the area CD EG a very material 


Fig. 66. 

It is this stage compression with intermediate cooling, and 
with spray injection also, which Professor Riedler has adopted 
in the new compressors for Paris, with marked increase of 
efficiency. Fig. 67 shows an actual diagram from one of the 
Riedler compressors, and it will be seen that the excess com- 
pression is only 1207 per cent., so that in this case ij^ has the 
high value 0-89. 

Fig. 68 gives the diagram of another Riedler compressor at 
Paris, running at a higher speed. 



In using air for transmitting power it is important, with a 
view of diminishing the size of the mains^ to adopt high initial 

Fig. 67. 

pressures, and it is the inefficiency of the compressing process in 
ordinary machines which has hitherto prevented the adoption of 
such pressures. Provided the efficiency of the compressors can 
be increased, much greater pressure loss can be permitted in 

« ^rKS PtttMlM: 

4C7 err n/t nfiTioim 

Fig. 68. 

the mains, and consequently much smaller mains will suffice to 
transmit a given amount of power. 



Professor Riedler and Professor GutermuiKs Experiments on 
the Efficiency of Compressors. — If, as above, Uj=i7i U is the 
work expended in compression, after deducting the friction of 
the mechanism, and 

Ua = 27,710 log. ^ 


is the useful work done in compression, then 

i7, = U2/u,=U2/t;iU 

is a coefficient of efficiency of the compressing process, which 
includes both the waste of work in useless heating and any 
losses due to clearance, valve resistance, &c. 

Professor Riedler has obtained a series of indicator diagrams 
from diiferent compressor cylinders. On these he has drawn an 
ideal isothermal diagram, without clearance or valve losses, and 
for compression to the pressure in the main. The excess of 
area of the actual diagram over the ideal diagram is the work 
wasted in the compressing process, and from this 17, is easily 
calculated. The following table gives a series of such results: — 


Type of compressor 

Pressure In 
main, jv, 

CoUadon, St. Gothard 



Slide- Valve Compressor 



Riedler Two-stage 

Lost work 

in per cent, 
of useful 



















The table shows how very low is the efficiency of some of 
the older compressors. Even in the compressors in Paris with 
single-stage compression i;j=0*70. Hence, if the friction of 
mechanism is taken account of, and rj^ is put at 0*85, the 
resultant efficiency of the compressor mechanism and the com- 
pressing process is 0*85 X 0*70 = 0-595, or less than six-tenths 
of the indicated work in the steam cylinder is usefully expended 
in compression. The Riedler two-stage compressor gives a 
much better result. Taking i7j=0-87, its efficiency is 0*77. 



Professor Riedler has given some other figui^s, not based, as 
the above are, on the measurement of single diagrams, from 
which the eflSciency of the Paris compressors can be calculated 
in a more trustworthy way. The following data are given as the 
result of a series of experiments with each compressor : — 

Type of compressors 

Pazman machine (Sturgeon compressors) 
GockeriU machine (Dubois - Francois 

Riedler machine (Two-stage com- 

For each indicated steam h.p. 

Cubic feet of air 
at atmoHplieric 
pressure com- 
pressed per tiour 



Pressure in main, ' 

lbs. per sq. inch 




From these figures the following are calculated :- 

steam i Weight of 
work ))er | air corn- 
hour in pressed in 
foot lbs. lbs. 

Paxman . . 1,980,000 
Cockerill . { — 
Biedler . : — 



Calculated .^"i^- 
, work per %^": 
work per lb. for *=**°*'> 
lb. of air, ! isothermal I ''» *»» 
=U t compre»- I _ V.^ 
1 Blon, XJ, ; x; 


97,830 ' 53,920 0-551 ; 085 ' 0648 
86,340 — 0-624 085 0-735 
70.550 — 0-764 . 087 . 0*898 

These results agree well with those obtained above in a 
different way. 

Action of the Air Motors, — An air motor is simply a reversed 

air compressor. Hitherto the conditions of eflSciency in air 

motors have received very little attention. In Paris many of 

those used are of small size, and in these a good efficiency is not 

to be expected. The best results have been obtained thus far by 

adapting old steam engines to work as air motors, and this can 

be done with very little trouble. It is specially desirable that in 

I an air motor the cylinder clearance should be small or the com- 

\ pression sufficient. Probably the greatest source of avoidable 

I waste in air motors has been leakage at the piston. In a steam 

j engine the condensation on the cylinder wall helps to render 

* the piston tight. In an ordinary air engine, the cylinder surface 


IS more or less dry, and the waste with air from leakage at even 
small apertures is very great. 

A considerable economy can be secured by re-heating the 
air in a simple form of stove before admitting it to the engine. 
At first sight this seems a complication likely to involve as 
mach trouble as a steam-boiler. That, however, is not at all the 
case. The re-heating appliances are simple, there is no risk of 
dangerous explosion, and the amount of heat which it is desirable 
to give to the air is insignificant compared with that required 
in raising steam. Professor Riedler tried an old 80-h.p. steam 
engine in Paris, which had been adapted to work as an air 
motor, and which was actually giving 72 indicated h.p., with 
compressed air at 5^ atmospheres pressure. It was using about 
31,000 cubic feet (reckoned at atmospheric pressure), or about 
2,376 pounds of air, per hour. This air was heated to a 
temperature of about 300° Fahrenheit by the expenditure of 
only 15 pounds of coke per hour. On favourable assumptions 
a steam engine working to the same power would have required 
ten times this consumption of fuel at least. Re-heating the air 
has the practical advantage of raising the temperature of exhaust 
of the motor, and for the amount of heat supplied the economy 
realised in the weight of air used is surprising. The reason of this 
is that the heat supplied to the air is used nearly five times as effi- 
ciently as an equal amount of heat employed in generating steam. 

In certain cases the air-motor cylinder has been jacketed 
by hot air. This increases again the amount of work obtained 
per pound of air used. It can hardly be considered a thermo- 
dynamically advantageous process, but it may have advantages 
practically in raising the exhaust temperature. Lately in some 
cases, water has been injected in small quantity into the air while 
passing through the re-heating stove. This passes into the 
engine as steam. It condenses during the expansion, yielding 
latent heat to the air, and thus raising the temperature of 
exhaust. Whether it is advisable from a purely thermodynamic 
point of view may be doubted, but it seems to have practical 
advantages possibly in lubricating the cylinder and preventing 
leakage. When steam is employed in this way the expansion 
curve rises above the adiabatic and becomes nearly an isothermal. 
The steam may amount to about 5 per cent, of the weight of 
air used. 


Useful Work done hy an Air Motor. — ^Let the air be delivered 
from the main to the air motor at the pressure Pj in lbs. per 
square foot, or p^ in lbs. per square inch. (1.) Let it be 
supposed that the air is used in the motor cold, its temperature 
being taken at 60"^ Fahrenheit and its volume in cubic feet per 
lb. being v^. Expanded adiabatically in an engine down to 
atmospheric pressure, p^^ the work done would be in foot lbs. 
per lb. of air. 

= 95,600 [l - ( ?''i y^n . . (16) 

The actual work obtained in any given motor using air cold 
will be less than this in consequence of incomplete expansion, 
valve resistance, clearance, leakage, and other losses. Let the 
actual work shown on the indicator diagram of the air motor be 
u^. Then if 

«73 is the efficiency with which the fluid is used in the particular 
air motor, a coeflBcient which must be determined by experi- 

(2.) Let it be supposed that the air, arriving in the main 
with the temperature T^ (absolute), is re-heated to a temperature 
T3 before being used in the motor. As it is re-heated at constant 
pressure, the amount of heat to be given to each lb. of air is 183 
(T3— Tj) foot lbs., or 0-2375 (t,— t,) thermal units. Then the 
work of adiabatic expansion from a pressure 2^2 ^ atmospheric 
pressure p^ will be — 


= 95,600 J.- [l-(^A)"'] . . (17) 

Generally the work shown by the indicator diagram of any 
engine will be less than this, for the causes mentioned above, 
and if 1/3 is the efficiency with which the motor uses the fluid, 
the indicated work will be — 



There is no reason for expecting 1/3 to be different in this case 
from what it was in the previous one, unless hot-air jacketing, 
or steam injection, is used. In that case 173 will be larger, and 
may be taken to include the additional work due to heat supplied 
•during the stroke. 

There is yet one more source of loss in the motor, the friction 
•of the mechanism. If r)^ is the efficiency of the mechanism, 
then — 

U5='^4 U3 or 17, u'4 

is the effective or brake work of the motor per lb. of air used. 

Expenments on Air Motors by Professor Eiedler and Professor 
Owtermuth, — In the following tables some of the experiments 
Are quoted, together with the work per lb. of air, and the values 
•of 1/3 and 7}^. When Professor Riedler does not give the 
indicated, but only the brake horse power of the motor, it will 
be assumed that 974 = 0*85 in order to calculate a probable value 
•of riy 

It will be seen that in the older small motors the efficiency 
with which the fluid is used ranges from 0*37 to 0*44, which 
perhaps, considering the kind of motor, is a good result. In 
the later machines, arranged to work expansively, the efficiency 
ranges from 0*58 to 0*87, results remarkably good for motors 
;S0 small. The coefficients are rather higher with re-heated 
-air, showing that the work done is increased even in a rather 
higher ratio than Tj/t^. 

Some other experiments on small motors may be passed over 
in order to consider some experiments on an old Farcot steam 
■engine with Corliss valves, which had been converted for use 
as an air motor. 

This engine was nominally of 80 h.p., and worked at 72 
i.h.p. in the trials. In all cases the air was re-heated before 
use to about 300° Fahr. The cylinder was also jacketed by the 
hot air on its way to the cylinder chest. 

The efficiency is therefore 0-81, an extremely good result. 

Of all the work obtainable by the expansion of 1 lb. of 
air received at 95*5 lbs. per square inch, and at a temperature 
of 300° Fahr., four-fifths is obtained as effective work on the 
brake. However, part of this work recovered is borrowed from 
the hot air before admission to the cylinder and given back 









00 *H IQ ^ 

90 lO ;d *H o CO 
30 o o to S t^ 

CO -♦ OO 



lO «o t^t*ooao 

ooo 666 



ta ta tQtoiQ to 

00 X 30 QO 30 00 


oo oooo 


G<1 -^ '«<< CO 


^ 1^-^ "^ 

open 00 CO 

'^t* »p CO «© t- b- 





t-» t-» o o 


<0 «0 — *G 

CO CO <M CO CO 1^ 

U3 io »o :o 

lO to CO 04 O) p^ 

CO 00 "^ -^ 

CO 00 « "* -^ ^ 



•^ -^ « I-. 


kO h- t- (M ^ ..H 

,-1 so iO Si 

CO »-^ -t O CO CO 







t* »o >o o 

CQ (M c: CO t« t» 

t» X b- -^ 
t- -^ <M C>1 

^ Si X C- CO S 




ra S ai S 




o «o »-• to 

O CO O f-t X X 
t« X CO 'A ^ ^ 1 

Air use 


per brak 

per he 


CO "* t^ c^ 

CO Ci « o 


-^^c^o^oaxx ' 

re, lbs. 


>Q IQ lO t.O 


lo lo »Q ioua »o 




t^ ^-'^- t>. 

t> !>. t« t^ C« t> 






OiO 1 -^ 

CO -^ "^ X IX 



1 t- 

CO ^»C CO 1 (N 

1 1 1 











lO «o c^ -♦• 

1-1 1-1 

rH ,-1 i-« 




«o o 

•S si. 

C QO ^ C^ 

•^ X o o >i c<i 


»0 O «0 -r 

OJ ^ CO « b- (M 

O r^O rH 

1-1 1-4 CI »-< ©« CO 


o ro 

CO 00 "* ?0 

^ r^ O O »o ^ 
c>i o o -H CO -^ 

^ S S3 


O r-4 3^ C^ 

CM CM C^ <N c^ c^ 




rifit Hjfl 


iH 1-1 

^ ^ CM i-i C4 04 


p lb 11 

ca o ^ t* 

X Od A X 

6 666 
















7_ e cj 





^ 1 


'S£ II 



O '^•cT "^ 
«o o «o to 




lip -«4 

> lb «b 


I. « »0 CO c 

^1 I cococ 

2 ;£. b- la I 

•O !r ^ b-C 




^ iO O X 




to it by the jacket. That the jacket considerably affected the 
working during expansion is shown by the temperature during 
exhaust. The following Table gives the temperatures the air 
would have reached by adiabatic expansion, and the actual 
temperatures of the exhaust in the experiments above : — 

Initial temperature 

Pinal temperature for 
adiabatic expansion 





+ 70 
+ 84 
+ 96 
+ 120 

Increase of 

of exhaust 
due to 








781 . 



Practical Calculations on the Distribution of Power hy Com- 
pressed Air. — It will be convenient first to give a summary of 
the formulas required in settling a system of compressed air 
distribution, and afterwards to discuss some special cases. 

Let P„ be the pressure in lbs. per sq. ft., p^^ the pressure in 
lbs. per sq. in., v„ the volume of a lb. in c. ft., 
Ta the absolute temperature of air admitted 
from the atmosphere to the compressor. 

Pp j?p V,, T, the same quantities for air discharged from 
the compressor into the main. 

Pj, ihf ^2' "^2 *^® same quantities for air arriving at the 
point of consumption in the main. 

u = indicated work done by the steam on the piston 
of the compressor reckoned in h.p. 

U, = t;, U = corresponding indicated work on the air in 
the compressor cylinder. 

Uj = 1/2 Uj ^ useful work of compression, or work of 
isothermal compression from p,^ to j^,. 

Uj =r available work of air arriving at the motor, or 
work of adiabatic expansion from p^ to pu. 

U^ = 1/3 Ug = indicated power of the air motor. 

Ug = ?/4 U4 = effective or brake work of the air motor. 
Taking p« = 2116-3, ;7, = 147, Va = 1309, t„ = G2r. 

p„v« = 27,710 (1) 

If w lbs. of air are compressed per second by u h.p. in 
the steam cylinder corresponding to Uj i.h.p. in the compressor 


cylinder, then, allowing both for friction of mechanisni, and 
clearance, leakage, and other losses in compression — 

550i7,i7,c=27,710wlog. ^'» 

w= - ''^^«'^-- . . . . (2> 

50-4 log. ^^ 


When, by conduction in the main, the air is again at its initial 
temperature T„ — 

PiVi = 27,710 

2>iVi = 192-3 . . . . (3> 

If r J is the initial velocity of the air in the main, the diameter 
of which in feet is d — 

'^(/^',=wvi = 192-3 w/j>, 

<?=15-64a/-^ . . . (4> 
V 2\vt 

The pressure falls in a main of length I, in feet, from p, to p,, 
the amount being given by the equation — 

i',_ ^ /1 1 _ ^'i*^ 1 rA 

p,~V 1 74,300,000(if • • ^"^ 

The available work of the air arriving at the motor with the 
pressure ^'j at atmospheric temperature T, is, if it is used with- 
out re-heating, in h.p. — 

= 173ow[l-(£^)"^] . . . (6) 
If re-heated to T, before admission to the motor — 

Indicated work of motor = u^ = rj^v^ ... (8) 
Brake or eflfective work of motor = u^ = ly^u^ . (9) 

„..=i73-5wr.[i-(r;)"'] . . (7) 

Values assumed for the Coefficients of Efficiency. — It appears 
that the eflBciency of the mechanism of compressors is from 0*85 



to 0*87. In the following calculations it will be assumed that 
T/j = 0*85 . As to the efficiency of the process of compression, 
this varies greatly with the type of compressor. In some of the 
older single-stage compressors it is as low as 0*5. But taking 
the best of those tried by Professor Riedler and slightly rounding 
off the values — 

For single-stage compressors . . . ^^=0*7 
„ two-stage „ ... 172 =0-9 

The loss in the main must be calculated for each special case. 

For the air motors, as it is not intended to discuss the use 
of air on a small scale, it will be assumed that t)^ = 0*85 and 
7f^ = 0-9. These values are slightly below those obtained in the 
experiments, and the results should therefore be such as are 
practically realisable in ordinary work. 

Case I. — 10,000 h.p. is to be transmitted a distance of 2 
miles. Taking ij^ = 0*85 and t)^ = 0*9 (two-stage compression) 
the useful work of compression will be 7,650 h.p. Now, let the- 
air be compressed to 4, 8, and 1 2 atmospheres on the gauge^ 
then the weight of air compressed will be as follows : — 

Case I 

Initial gauge 




Initial pressure, 
lbs. per sc]unrc i_' 

inch abmlutc I Pa 

Weiglit of air 
' compresserl In lbs. 
I per second ca W 





To ascertain suitable diameters for the main, let initial 
velocities of 30, 50, and 75 feet per second be assumed. For 
these velocities, and in the cases described, the diameters re- 
quired will be as follows : — 

Case I 

> Initial 
i lbs. per sq. in. 
I absolute 


velocity in 

the main, 

ft. i)er secoiul 

Diameter of main 

In feet 























In inches | 

39 I 

30 I 

25 [ 

19 I 

16 I 

19 ; 




None of tbese mains are impracticable in size, and for the 
higher pressures and velocities they are surprisingly small, con- 
sidering that they are shown to be capable of transmitting 
10,000 indicated steam h.p. a distance of 2 miles. It remains 
to examine whether the loss of pressure is serious. 

Case I 

lbs, per sq. in. 
I absolute/;^ 

Telocity in 
the main, 


Dlameter ; I^^„ 

of the main RJ*"?'*" 

in feet ^"^i^"^' 

Loss of pres- 
sure in per • 
cent, of initial 


1.32-3 i 







• 60 


















\ 50 










None of these losses would render the utilisation of the power 
impossible, and it is only at the highest pressures and velocities 
that the fall of pressure is serious. 

Lastly, the power developed at the air motor due to 10,000 
indicated steam h.p. in the compressor has to be calculated. 
The air being delivered at a distance of 2 miles, and used cold, 
the h.p. obtained will be as follows : — 

Case I 

b . 

Initial I 
prt^, I 
lbs, |>er sq. [ 
iu. ubsoluLe 


132-3 ' 


Terminal I 
prepare in ' 

the main, ' 
lb& per sq. 
in. abiM>lute 




f 128-3 






velocity in 
the main, 

fwt iwr 




h.p. of air 


I Brake or | 
; effective h.p. 
i of air motor 








, 4.755 












It is very striking how little the efficiency is affected by 
considerable changes in the initial pressure and velocity in the 
mains ; with the exception of two cases, for 10,000 steam indi- 
cated h.p. expended at the compressor there is obtained from 
4,400 to 5,100 indicated h.p. at the air motor. That is to say, 



the efficiency of the whole arrangement, compressor, main and 
air motor, when the air is used cold, ranges from 44 to 51 per 
cent., and that with mains of quite moderate size. 

If the method of re-heating at the motor is resorted to, 
which in no case involves any great additional trouble, and which 
from the small amount of fuel required can oflen be carried out 
with very little additional expense, the h.p. obtained will be as 
follows. Let it be supposed that the air arriving at the motor at 
60° Fahrenheit is re-heated to 300° Fahrenheit, as in the case of 
the Farcot steam engine, details of which are given above. 

Ta 761 1 .^ 

T« 521 

Case I 




lbs. per sq. 




lbs. per aq. 


Initial vclo- 
oity in the 
main, feet 
per second 

Indicated Brake h.p. 
h.p. of motor . of motor 

r 720 



73-5 1 









( 128-3 



132-3 1 \ 










191-1 , ] 









Here again, excepting two cases, the efficiency reckoned on 
the indicated power is from 64 to 75 per cent., neglecting the 
cost of the fuel for re-heating. 

Case II. — A long-distance transmission may now be dis- 
cussed. Suppose, as before, that 10,000 i.h.p. is developed 
in the steam cylinders of the compressors, and is to be trans- 
mitted a distance of 20 miles. Taking the same pressures 
as before, the calculations of the weight of air compressed 
need not be repeated. The initial velocities previously assumed 
will, however, be excessive, because the velocity in the main 
increases as the pressure falls, and in this case, as the fall will 
be considerable, the terminal velocities are much greater than 
the initial velocities. Initial velocities of 20, 35, and 50 feet 
per second will be as^sumed. 




Caw II 

Ibfl. per Bqnare 
Inch alMolute 


Initial Telocity, ■ 

Diameter of 

Diameter of 

feet per second 1 

main in feet 

main in inches 




























None of these mains are impracticable in size. The follow- 
ing table gives the calculation of the pressure loss : — 


Loss of pres- 1 

Case II 



Diameter of 


sure in per 1 



main in feet 

lbs. per sq. 

cent, of initial- 

1 inch 

pressure 1 




d , , , 





1 47-6 





— 1 



1 116-5 


e , . . 





' 12-6 










/ . . . 




I Impossible 





~ i 

It turns out on calculation that some of the cases assumed 
are impossible. That is, the whole initial pressure is insufficient 
to give the assumed initial velocity. In one other case, the 1-91 
foot main, with a velocity of 35 feet, the loss of pressure in the 
main is impracticably large. There remain, however, four casbs 
in which neither the size of main nor the loss of pressure in a 
transmission to a distance of 20 miles is such as to render the 
transmission impracticable. In three cases the loss is remark- 
ably small, and the terminal pressure quite suitable for applica- 
tion. For instance, it appears that the air compressed by 10,000 
h.p. to 132'3 lbs. per square inch can be transmitted to a distance 
of 20 miles in a 30-inch main with a loss of pressure of only 12 
per cent. 

The power delivered at a distance of 20 miles by air motors 


using the compressed air can now be calcalated. If the air is 
used cold, the power obtained will be as follows : — 




lbs. per sq. 

in. absolute 



Terminal I Initial I 

pre^ure, ' velocity in Indicated h.p. 
lbs. per sq. I the main, ft. of motor 
in. absolute I per second 


' 68-0 






Brake or 

effective h.p. 

of motor 


If the air is re-heated to 300° Fahrenheit, as assumed in the 
previous case — 





lbs. per sq. 




lbs. per sq. 



velocity in 

main, ft. per 


Indicated h.p. 
of motor 

Brake or 

effective h.p. 

of motor 



r 680 

t 47-6 






Here the eflSciency of the whole arrangement — calculated on 
the indicated power, the air being delivered at a distance of 20 
miles, and including all losses — is 40 to 50 per cent, if the air 
is used cold, and 59 to 73 per cent, if the air is re-heated. The 
results are based absolutely on efficiencies already obtained in 
similar cases, and the sole loss neglected is possible leakage in 
the mains. 

Fig. 69 is drawn to scale for Case lie. It is a diagram 
showing the relation of the work expended and useful work 
recovered when 10,000 h.p. is transmitted 20 miles in a main 
30 inches in diameter. E B G R T D is the work expended in a 
two-stage compressor, compressing 1 lb. of air from 14*7 to 
132 '3 lbs. per square inch. By cooling in the mains the 
volume of the air shrinks from B G to B c. The frictional resist- 
ance reduces the pressure to 116*5 lbs. per square inch, when 
the air has the volume MK determined by the isothermal. 
The work done by an air motor using the air cold is M k A E. 
If the air is re -heated to 300° the volume expands from M K to 




ML (1-46 times). Then the work of the air motor, using 
re-heated air, is mlhe, and klha is the important gain of 
work due to re-heating. Of course the areas given by the 
diagram have to be multiplied by the efficiencies given above. 

116-5 Lbs a j^vvjaa t^^v 

! 1309 Cu? FT T 

Fig. 69. —Combined Action of Compressor Main and Motor. 

In an air main the air expands as the pressure falls. Hence 
in a very long'main the diameter should increase as the pressure 
falls. An expanding main of this kind has not actually been 




Generally, boilers producing steam are adjacent to the engines 
generating power. In special cases it has been necessary to 
convey the steam not inconsiderable distances before it was 
used in engines. For instance, underground pumping machinery 
in mines has been worked from steam boilers at the surface by 
steam conveyed in pipes, protected as far as possible from 
heat losses. Nevertheless^ it would hardly have been thought 
reasonable to distribute steam widely to considerable distances 
for power purposes, but for a secondary object. Steam dis- 
tributed firom a central station through pipes can be very con- 
veniently used for heating purposes, as well as for power 
purposes. The defects of steam, as a means of distributing 
power, may be balanced by its advantages as a means of distri- 
buting heat. At any rate, in the United States the experiment 
of distributing heat and power from a central station, by steam, 
has been tried on a very large scale, and .with a considerable 
amount of success. 

In 1877, Birdsill Holly patented a system of steam distri- 
bution for heating purposes only. The steam was conveyed in 
pipes, having anchored stuffing boxes at distances of about 
1 00 feet, so that the expansion and contraction of the pipes was 
provided for. Radiation was diminished by covering the pipes 
with asbestos and wood. The steam was delivered into the 
houses through a reducing valve, and used in ordinary heating 
coils. The condensed steam was discharged through steam traps 
into the sewers. Generally, the condensed steam, before being 
discharged, was taken through coils in a chamber, through which 
fresh air entered the building. The air was warmed, and the 
condensed steam reduced in temperature to about 100'^ Fahr. 


The heat was thus economised, and the condensed steam dis- 
charged at a temperature at which it caused no inconvenience. 
Generating and distributing plants on the Holly system were 
erected in many towns. In Lockport, for instance, in 1879, the 
main steam pipes extended a distance of 16,000 feet. 

As early as 1869, Dr. C. E. Emery investigated the problem 
of distributing steam in New York.^ His studies led to the 
creation of the largest system of steam distribution hitherto 
carried out. Dr. Emery concluded that steam could be eco- 
nomically distributed, from one station, to buildings within a 
radius of half a mile. Ten plots of land for stations in New 
York were secured, and the construction of the works com- 
menced in 1881. Up to the present two steam stations have 
been put in operation, a down-town station, termed Station B, 
with boilers working to 16,000 h.p., and an up-town station in 
58th Street, designed for boilers of 3,000 h.p., and having about 
half this power at work. 

The pipes from the down-town station extend through 
5^ miles of streets; those from the up-town station through 
2^ miles of streets. 

Bmvn-iown Station, — This was erected on an irregular plot 
about 75 feet by 120 feet. To obtain space for the boilers, they 
are arranged in four tiers, each tier in a separate storey 20 feet 
in height. There is a fifth storey for coal storage, and a base- 
ment for miscellaneous purposes. Each floor is arranged for six- 
teen boilers of 250 h.p. each, placed in two rows, facing a central 
charging floor. The chimneys are near the centre of the space. 
The coal is dumped into small cars in the basement and then 
lifted to the top storey, where it is discharged into coal bins. 
The coal descends by gravity in chutes to the several floors. 
The ashes are discharged down chutes to the basement. The 
boUers are Babcock and Wilcox boilers. Fig. 70 shows generally 
the arrangement of the building, with one boiler indicated in 
place. At this station the pressure is generally maintained at 

* * District Distribution of Steam in the United States,' Proc. Inst, C, -EL. 
vol. xcvii.; *The Station B Chimney of the New York Steam Company/ 
TraTU. Am. Soc, C. H., 1885; 'District Steam Systems/ Trans. Am. Soe. C. £., 
1891 ; * The Comparative Value of Steam and Hot Water for Transmitting 
Heat and Power/ Trans. Am. Soc. Mech. E.y vol. viii. All these papers are by 
C. E. Emery, Ph.D., of New York, and they contain nearly all the information 
available on the subject of steam distribution. 



80 lbs. per square inch. Down to 1887, 48 boilers were installed 

in three storeys, the fourth 

storey being used for coal 

storage. The fifth storey 

had not been built. In 

1891 it appears that four 

more boilers had been 

placed in the fourth 


Up'toum Station, — 
This is designed to con- 
tain twelve boilers of 200 
to 250 h.p. each, of which 
half were in place in 
1891. The boilers are 
on one floor, with coal 
storage above and base- 
ment below. The plant 
was designed to carry 
steam of 80 lbs. pressure, 
but the pressure has 
generally been lower, and 
the steam is supplied in 
winter only. 

The steam distributed 
is used chiefly for heating 
purposes, but a consider- 
able quantity is also sup- 
plied fix)m Station B for 
power purposes. It is 
used for driving steam 
engines working printing 
presses, electric lighting 
machinery, lifts and ven- 
tilating machines. In 
1887, about 500 steam - 
engines were supplied 
from Station B. 

Gust of Steam, — The 
<;harge for steam is based Fig, 70. 

se/tte or retr 


on the quantity of heat snpplied. The unit of heat is a ' kal,' 
which is defined as the heat required to evaporate 1 lb. of water 
from 100° at 316°, or at 70 lbs. pressure per square inch. One 
kal is therefore about 1,110 thermal units. Mr. Emery reckons 
that, on the average, 30 kals per hour give one i.h.p., when the 
steam is used in ordinary small engines. 

When the steam is used for heating only, a rent for the 
steam supply is paid on an estimate of the heat required, based 
on a survey of the building. This method has not proved 
altogether 8atisfa<5tory. The charge for steam for power pur- 
poses, based on the nominal power of the engine, has also proved 
unsatisfactory. A meter is now used in many cases, which 
gives a graphic record ^ of the amount of steam delivered, and 
although not entirely satisfactory, this appears to have been fairly 

The ruling price for steam at first was 50 cents per 1,000 
kals, which, at 30 kals per i.h.p. per hour, is equivalent to |ti. 
per i.h.p. hour. It is stated that now a sliding scale is adopted, 
the charge being 70 cents per 1,000 kals to small users, and 
40 cents to large users. These charges are equivalent to 1 21. lOs, 
and 71. lOs. per h.p. per year of 3,000 hours. To newspapers 
and other establishments working chiefly at night the charge 
is 30 cents per 1,000 kals, equivalent to 5Z. 12s. per year 
per h.p. 

WorJii7ig of the System. — The steam mains are stated to be 
in as good condition as when laid, though they have been under 
a continuous steam pressure of 80 lbs. per square inch night and 
day. On the other hand, the condensed water return mains 
have proved a failure from external corrosion. 

To a great extent they are now disused. Mr. Emery appears 
to think that the temperature of the steam mains kept them ex- 
ternally dry, but that the return mains, at a temperature of 
200° to 212°, did not keep dry, and the corrosion was specially 
active at that temperature. (Emery, ' District Steam Systems,* 
p. 193.) With the earlier Holly system, worked at a lower 
, pressure and in smaller towns, it was possible to cool down the 
exhaust steam to a temperature at which it could be discharged 
direct into the sewers without mischief. In the larger system 

A curve is drawn on a paper strip, the abscissse representing time and the 
ordinates the weight of steam flowlDg. 


in New York, where in some cases the steam is supplied to con- 
sumers having no control of the basements of the buildings, and 
where the steam pipes themselves must be drained at every dip, 
it is more difficult to ensure the discharge of the water at a low 
temperature. If the water is discharged at a temperature above 
212°, it generates steam when the pressure is released. Mr. 
Emery states that since the use of return mains was discontinued 
much steam is finding its way into the sewers, and escaping at 
times from man-holes. Mr. Emery's present opinion is, that 
rolled brass return mains should be used, with cast-iron fittings. 
If the feed is returned to the boilers at 200° or 212°, he estimates 
that there is a saving of 10 per cent, of fuel. This would make 
it pay to use brass return mains. Mr. Emery believes that the 
steam system has been a great public convenience, and that, 
where property owners would associate themselves to erect a 
steam plant for the improvement of their property, the enter- 
prise would be a remunerative one. He also thinks that steam 
systems for winter use only, the plant being shut down in 
summer, would pay well, partly because very cheap coal can be 
used. The New York plant, in spite of some faults almost 
inevitable in a new enterprise, is running, and yielding a large 
income. Many buildings would have their rental value materially 
lowered if the steam supply were cut off. 

The Pipe Maiiis, — Cast-iron pipes with flanged joints have 
been used, but in the United States the pipes are welded 
wrought-iron pipes. In the original Holly system the joints 
were stuffing-box joints, to permit free expansion and con- 
traction of the pipes. The stuffing boxes give great trouble, 
even with low pressures, and involve the construction of a great 
number of man-holes for access to the joints. In the New 
York supply thick cast-iron flanges are fixed on the ends of 
the wrought-iron pipes. At first the wrought-iron pipe was 
fixed in a slightly conical hole in the cast-iron flange by rolling; 
the end of the pipe was turned square and abutted against 
a turned shoulder. A caulking space a quarter inch wide 
and one inch deep was left at the back of the flange. The 
later pipes used have screwed ends. A cast-iron flange is 
screwed on, and iron cement caulked into a space at the back 
of the flange. Some of the joints have flat faces on the flanges ; 
others have a spherical joint on the flanges, permitting a 



limited variation of direction of the pipes. The joints are 
made tight by a copper ring smeared with thin red-lead. 

At New York the large pipes are 16, 15, 13, and 11 inches 
external diameter and about a quarter inch thick. The smaller 
pipes are the ordinary standard wrought-iron piping. For the 
larger pipes, if p = steam pressure in lbs. per sq. in., d = dia- 
meter in inches, t s= thickness in inches, then 

t =z cp d 

where c is 0-00022 to 000026. 

Variators or Expansion Joinis. — The variation of tempera- 
ture in steam pipes is very great, and provision must be made 

Fig. 71. 

Fig. 72. 

for permitting expansion and contraction withont straining 
the joints. If the range of temperature amounts to 130° or 
150° the pipes will expand and contract -^^^th to y^th of 
their length, or, say, 1 inch in 50 to 62 feet of length. Holly 
used stuflSng boxes, but these give great trouble from leakage 
and the wearing of the packing by the sliding of the pipes. 
Bends (fig. 71) may be used, but these allow only limited 
motion. A better plan is to combine bends and stuflBng boxes 
(fig. 72). Then the pipe rotates in each stuffing box without 
sliding longitudinally, and the wear of the packing is diminished. 
Flexible diaphragm joints have been tried, but as usually made 
they permit only a very small amount of motion. If the dia- 
phragm is strong enough to resist the pressure, it is so stiflF that 



only a small amount of bending is possible without overatraining 
the diaphragm. 

In the New York supply Dr. Emery has adopted very thin 
diaphragm plates of copper, supported by radiating hinged bars. 
Thus a flexible and sensitive diaphragm is obtained, perfectly 
«team-tight, and the steam pressure is carried by the movable 
radiating bars. 

Fig. 73 shows one of these variators. The corrugated 
copper diaphragm is 004 inch thick. The inner and outer 

Fig. 73. 

edges of the diaphragm are flat, and are clamped between strong 
cast-iron plates. On either side of the diaphragm are the loose 
bars resting at each end on rough knife edges. Each radiating 
plate supports a part of the diaphragm, without in any way 
hindering its motion. The length of the radiating plates is 
6 inches. The set on one side of the diaphragm resist the steam 
pressure ; the set on the other side support it, if a vacuum is 
accidentally formed in the pipes. The movement at each 
variator may amount to 1 J or 1^ inch without overstraining 
the diaphragm. A variator is provided for each 50 feet of pipe. 


Sometimes for convenience two diaphragms are arranged at one 
variator. At the elbows and at the variators the pipes are 
anchored to the brickwork. The service pipes are taken off 
from the anchored ends of the pipes. 

Protection agahist Radiation Loss. — To prevent radiation 
from steam pipes various materials are used — straw laid parallel 
to the pipes and covered with loam, felt, cork, fossil meal, pajyier 
mdchs, slag wool, and asbestos. Slag wool is very effective, and 
is obtained by blowing steam through molten slag. There 
seems to be some doubt, however, whether, if the outside of 
the pipe is moist, it does not increase corrosion. In any case 
it is desirable to keep the outside of the pipes dry, and this is 
most likely to be secured if they are kept continuously heated. 

Let k lbs. of steam be condensed by radiation per sq. ft. of 
the surface of the pipe per hour. Then if Gj is the weight of 
steam entering and Gj the weight of steam delivered at the 
other end of a main of diameter d and length I 

The fraction of the steam condensed and for useful purposes 
wasted by radiation is 

_ Gj — Gj _ irdlk 
^ G^ "" Gj + Trdik 

Obviously the proportion of the steam wasted increases as G,, 
the steam used, diminishes. It will be greater as the demand 
for steam is smaller. 

In New York the pipes are laid in brick trenches, the brick- 
work being kept 4 inches away from the pipes. The pipes are 
asphalted and the space between pipe and brickwork packed 
with slag wool. Over the pipes is a roof of short tarred planks 
bedded in cement. These are covered by tarred paper carried 
•well down the side walls to exclude percolating water. There 
are usually two intermediate piera with saddles supporting the 
pipe, in each 50-foot length between the variators. At all dips 
traps are connected for discharging condensed steam. 

Leakage. — It is clear that very considerable difficulties arose 
in New York from leakage of steam, and repair was difficult, 
because the steam supply could not be stopped. Dr. Emery 
attributes these difficulties entirely to bad workmanship. 


In 1886, the losses due to leakage were investigated, the 
return mains being then in operation. With steam for 6,000 h.p. 
supplied in winter and for 3,000 h.p. in summer it was estimated 
that the radiation loss amounted to 150 h.p. and the leakage 
loss to 500 h.p. That is 2^ per cent, and 8 J per cent, on the 
winter supply and 5 per cent, and 16^ per cent, on the summer 
supply respectively. In 1887, with nearly double as much steam 
supplied, the radiation loss was estimated at 350 h.p. and the 
leakage loss at 720 h.p. 

Size of Mains, — The pipes are designed for a velocity of 
.80 feet per second, and it is believed that at this velocity the 
pressure loss does not exceed 10 lbs. per sq. in. per half mile of 

Let w = weight of steam flowing per hour in lbs. 
d = diameter of main in inches. 

Then, Dr. Emery's rule can be put in the form, — 

w = 100 (fi 

Quantity of Power transmitted hy Steam Mains, — Let u be the 
velocity of the steam in the mains, and v the volume of a pound 
of steam in cubic feet. Then 

4 V 

If w = 80 feet per second, which is not excessive considering 
the small density of steam, 

w = c "j: d« 

where c is a constant depending on the steam pressure. 

Gauge preagure, 
lbs. per square inch 

Absolute pressure, 
lbs. per square inch 

46 i 60 

75 90 

Cubic feet per lb. 

7-04 11-36 

4-81 , 16-63 

126 140 318 I 26-15 

If 80 lbs. of steam per hour will develop in ordinary engines 
one i.h.p., then the h.p. transmitted by a main is 

H.P. = 120w= 94ccP 



Diameter of main 

Steam gauge 


Ibe. per square inch l 

Weight of steam 

lbs. per aecond 

H.P. transmittMl , 
by main 




267 . 




391 1 






45 1 




75 1 




125 1 












125 1 



These figures are necessarily rough values. But they show- 
that, with steam, very large amounts of power can be transmitted, 
without serious pressure loss, through mains of moderate size. 




The distribution of gas in town districts to many consumers for 
use in generating power involves nothing new or untried. The 
convenience and cheapness of this method of distributing a means 
of obtaining power are so remarkable, that a considerable 
development of the use of gas for power purposes is likely to be 
effected. Power can be obtained from a gas distribution in large 
or small quantities, with freedom from many of the drawbacks 
attending the use of steam power, and at a cost proportional to 
the amount of power actually used. Two independent systems 
of supplying gas for power purposes have to be considered. In 
one ordinary lighting gas is used, and the demand for gas for 
generating power is in that case an important secondary source 
of revenue for existing gas companies. In the other a gas of a 
cheaper description is manufactured and distributed specially 
for generating power. 

In the United Kingdom there has been expended on capital 
account in gas undertakings, chiefly for lighting purposes, a sum 
of 60,000,000Z. The amount of coal carbonised annually is 
10,000,000 tons, and the quantity of gas manufactured is 
98,000,000,000 cubic feet. If all this were used for producing 
power it would furnish about 1,000,000 h.p. for 3,000 working 
hours in the year. Lighting gas is already in many towns used 
on a considerable scale for heating and power purposes. Mr. 
Trewby, the President of the Institute of Gas Engineers, estimates 
that in the London district alone there are 70,000 gas cooking 
and heating stoves, and 2,500 gas engines. 

Gas has one advantage over electricity or compressed air^ 
namely, that storage can be so cheaply provided that the manu- 
facture can be carried on continuously and uniformly throughout 
the twenty-four hours. 


Use of Lujhting Gas for Qeneratin^ Power, — ^The cost of 
ordinary lighting gas is not so great as to preclude its use on a 
large scale for power purposes. Taking the cost of lighting gas 
at 28. to 38. per 1,000 cubic feet, and the consumption of gas in an 
engine with an ordinary varying load at 26^ cubic feet per 
effective h.p. hour, the cost of the power for gas only is 8Z. to 
12Z. per effective h.p. per year of 3,000 working hours. For 
interest on the cost of gas engine and engine house, depreciation 
and wages of engineer, an allowance of 3i. per annum per 
effective h.p. is sufficient. Hence the total cost of power derived 
from lighting gas would be from 111, to \bl. per h.p. per year of 
3,000 hours. 

It must be pointed out, however, that the price charged for 
lighting gas includes interest on a large network of mains, and 
the loss due to leakage over an extensive area. This part of the 
cost is not fairly chargeable against gas used for power purposes. 
In a distribution of gas for power only, the system of mains 
would be comparatively simple, and it would not be necessary 
to provide for the wide fluctuations of demand which occur in a 
distribution for lighting. There would be comparatively few 
consumers each taking comparatively large quantities of gas. 
It appears that the cost of manufacturing gas, including coal, 
wages, and petty stores, is about lOtZ. per 1,000 cubic feet. 
Probably 18rf. per 1,000 cubic feet would allow margin enough 
for profit and cost of distribution, to power users in a manu- 
facturing quarter not unfavourably distant from a generating 
station. But, at that price, the cost of power for gas only would 
be 6i. per effective h.p. year, and the total cost 9Z. per effective 
h.p. year, including interest and depreciation on the cost of gas 
engine and labour. 

The Bei^saiL Central Station for Electric Lighting. — This 
station was put in operation in 1886, having been erected by 
the Gas Company with the object of increasing the consumption 
of gas during the daytime, and at the same time of meeting the 
demand for electric light. It has since been extended, and it 
has been successful enough, both mechanically and financially, 
to show that the production of power from lighting gas on a 
fairly large scale is practically and commercially possible. 
The motor installation consisted at first of two 2-cylinder 
Otto system gas engines, of 60 h.p. each, one single-cylinder 


engine of 30 h.p., and one7of 8 h.p. The group of engines 
worked up to about 160 effective h.p. The dynamos were 
driven by belting and counter-shafts. In 1 89 1 , an engine of 1 20 
effective h.p. was erected with directly coupled dynamo, and one 
of the 60 h.p. engines and the 30 and 8 h.p. engines were re- 
moved. The jacket water is cooled by air coolers and used over 
again. The air coolers have 1,076 sq. ft. of cooling surface, 
an iiijector worked by pressure water from the town mains being 
used to circulate air through the cooler. The consumption of 
water is 5 gallons per h.p. hour for all purposes. 

In an electric station, gas engines have the advantage that 
they can be started and stopped when required, and they have 
no stand-by losses like those of steam boilers. On the other 
hand they do not work efficiently except at full load. At 
Dessau, a large accumulator battery is used for storing energy 
when the demand would not keep the engines fully loaded, and 
re-storing it in hours of small demand when the engines are 
stopped. The efficiency of the battery is 79 per cent, on the 
average of the year. About 52 per cent, of the whole supply 
passes through the battery, so that the waste of current in the 
battery is about 1 1 per cent, of the total yearly supply. 

The average gas consumption (before 1891) was 2G^ c. ft, 
per effective h.p. hour. * Motors of varying power were adopted at 
first, with an idea that they would best supply a varying demand. 
The constructors of the station now think that the accumulator 
battery renders this unnecessary and that motors of a larger 
and uniform size would be more economical. They claim as 
advantages of a gas plant compared with a steam plant that less 
space and less water are required ; that there is absence of smoke 
and danger of explosion ; and that gas stations can be distri- 
buted more easily over the district to be supplied. 

Disirihidion of Natural Gas at Fittshurgh, U,S,A. — A re- 
markable case of distribution of gas, for heating and power 
purposes, has been in operation at Pittsburgh.' The natural gas 
has almost entirely taken the place of coal in manufactories and 
for domestic heating, in a district where coal is exceedingly 

* It is stated to be now 39 cubic feet per EUowatt hour at the terminals 
of the dynamo. 

* See a paper by Mr. Andrew Carnegie on * Natural Gas,' read before the 
Iron and Steel Institute. 



cheap. Coal can be obtained at Pittsburgh for 4s. to 55. a ton, 
and coal slack at 28, to 28. 6d. a ton. 

The natural gas was met with in boring for oil, and was first 
used to raise steam for the oil pumping engines. At 18 miles 
from Pittsburgh an enormous outburst of gas occurred, which for 
five years was allowed to bum to waste. Then a company 
engaged to take it a distance of 9 miles to Messrs. Carnegie's 
works. They were to be paid for the gas the value of its 
equivalent in coal until the capital cost of the pipes was repaid. 
After that the gas was to be supplied at half the cost of its 
equivalent in coal. In 18 months the cost of the pipes was re- 
paid, and the gas was then supplied at half the cost of its equiv- 
alent in coal. It was afterwards conveyed into Pittsburgh and 
to still greater distances. When Mr. Carnegie described the 
operations, there were eleven gas mains, of 6 to 12 inches 
diameter, conveying gas to Pittsburgh. 

The largest well discharged 30 million cubic feet per day, 
and other wells half that quantity. At the wells the gas had a 
pressure of 200 lbs. per sq. in., and at ^lessrs. Carnegie's works, 
9 milt's distant, the pressure was 75 lbs. per sq. in. This gave 
rise to difficulties from leakage, and it was found desirable to 
reduce the pressure in the pipes, in towns, and even to place venti- 
lating pipes at every joint in the mains, leading the leakage above 
the level of street lamps. In using natural gas, one fireman can 
manage boilers developing 1,500 h.p. 

MannfaHure of Special Gus for Heatimj and Generating Fover. 
A cheaper gas than lighting gas can be manufactured for heat- 
ing and power purposes. (1) Frtnhicer (?<:«?, obtained by forcing 
air through incandescent coke or anthracite. The resulting gas 
is mainly carbonic oxide and nitrogen. (2) So called Waier Gas, 
obtained by injecting steam through incandescent coke or 
anthracite. Such gas has a volume of about 26 cubic feet to 
the pound, and will develop about 7,373 Th. U. per lb. In 
manufacturing water gas, air is first blown through the fuel till 
it is incandescent, and then steam, the alternation being repeated 
as necessary. (3) Bowson Gcis, made by passing air and steam 
through incandescent coal or coke. This gas contains hydrogen, 
carbonic oxide, and a considerable quantity of nitrogen. Four 
volumes of it are about equal in calorific value to one volume of 
lighting gas. It develops 160 Th. U. per cubic foot, or about 



2,382 Th. U. per pound. With anthracite at 135. a ton, Dow- 
son gas costs about 2d, per 1,000 c. ft. for fuel used, and exclusive 
of interest and depreciation of plant. Dr. Monaco gives the 
total cost of producing Dowson gas at 4d. per 1,000 c. ft., and 
its calorific power as one-fourth that of lighting gas. In that 
case Dowson gas is in heat value equivalent to lighting gas at 
Is. 4d. per 1,000 c. ft. (4) Mr. Thwaite has proposed for 
power purposes a gas of about 12 candle power, obtained by 
mixing lighting gas and producer gas. According to him such 
a gas could be manufactured for 4id. per 1,000 c. ft., and dis- 
tributed and sold at Is, 4d!. per 1,000 c. ft. Its calorific power is 
little less than that of lighting gas. 

The following table contains data of the density and calorific 
value of various kinds of gas : — 

Density and Calorific Value op Gas 


Calorlflo value 



.of products 
1 of com- 
1 buijtion 

Cubic feet 
per lb. 

for com- 
lier 0. ft., 
in c. ft. 


for com- 
per lb., 
in lbs. 

Per 0. ft. 

Per lb. 

1 with air 
: in c. ft. 
1 per eft. 

Manchester gas 


1 6-29 

American gas 






1 — 

London gas . 

' 3V7 





1 — 

Petroleum . 






Pittsburgh gas 







London gas . 






1 — 

Dows ^n gas . 






' 2-74 

Water gas (a) 

1 26-0 






M .. W 

' 20 8 






Dowson gas . 






1 200 


Not carbui 





Formula for Flow of G<is in Pipes, — Let p,, Pg be the initial 
and terminal pressures in a main of length L (foot units). The 
velocity of flow is given by tlie equation * — 

__ /\gcTm P^^ — P.,^] 

where u^ is the velocity at the inlet of the pipe. For pipes of 
circular section and diameter d, m = c^/4. For lighting gas 

Distribution of Power by Compressed Air/ Proo. Inst, C. K., 

* Unwin; 
vol. cv. 

8 2 


c =s 130 ; for Dowson gas c = 64. Let the temperattire be 
60® F., or the absolute temperature T = 521°. Then ct = 
67,730 for lighting gas, and = 33,344 for Dowson gas. f, the 
coefficient of friction, = 0003. Introducing the numerical 
quantities, for lighting gas — 

t*,= y^[ 181,700,000^ P'"/^' ] 

for Dowson gas — 

«.= y/ (89,450,000 ^- ^'* -/«'} 

where the pressures are in lbs. per square inch. 

When the initial velocity of flow is given, and the terminal 
pressure is required in terms of the initial pressure, for lighting 

2^2=i^iy^ |1 - 181,700,000 d) 
for Dowson gas — 

^^"^^Vr "897450,000 dl 

Case I. — In an ordinary gas distribution the difference of 
pressure producing flow is small, being about 2^ inches of 
water. If jPj = 14*7 lbs. per square inch, p^ = 14*7361, and 

^^f?" =0-00506. 
The equations reduce to — 

«j = 959-4a/'^ 

for lighting gas, and 

= 673-2^ J- 

for Dowson gas. 

The quantity of gas delivered in cubic feet per hour will be 

3,600 X jcPm,. 

Assume a distribution to a distance of 5,000 feet. Then, 
with the given pressure difference of 2^ inches of water column, 
the quantity of gas discharged and its equivalent in power will 
be as follows : — 



FowBB Transmitted in Oas Mains 
Length, 5,000 feet ; pressure producing flow, 2J inches of water column. 

A. — Lighting €Uu 

Diameter of main ' ^^j!fL\f;?^*^ 

Cubic feet 

of gas dellrered 

per hour 







B. — Dojvson Oas 

Diameter of main 
in inches 

Initial velocity 

in main, 
feettper seconci 

Cubic feet 

of gas delivered 

per hour 













H.P. at 26-6 

cubic feet per 

h.p. hour 





H.P. at 90 

cubic feet per 

h.p. hour 



It will be seen that none of the velocities under this pressure 
are excessive. The Dowson gas being heavier, the friction is 
greater and the quantity flowing is less. Further, as the heat 
value of Dowson gas is less than that of lighting gas, the amount 
of power transmitted in a main of a given size is only about 
one-fifth as much for Dowson gas as for lighting gas. 

Case II. — It may next be inquired what would be the result 
of using greater pressure to force the gas through the mains 
than is usual in supply for lighting. In ordinary gas mains it 
is found unadvisable to increase the pressure, because of the 
increase of leakage. In the distribution of gas for power pur- 
poses this objection would have less weight. The network of 
mains would be simpler, and, the consumers being fewer, there 
would be fewer joints and valves to cause leakage. By the 
adoption of some really efficient joint, like that used in the 
Paris air mains, leakage could be almost reduced to zero. The 
pressure which would, then, seem to be desirable for a gas power 
distribution ife the pressure which would produce in the mains 
the highest desirable velocity of flow. It may be taken from 
the analogy of compressed air mains that 45 feet per second 
is a quite unobjectionable velocity. 

Assuming this velocity as the initial velocity in the mains, 



the problem is to find the necessary initial pressure. The 
equations become — 

V 1 89,730 d) 
for lighting gas, and — 

Pi = - 



L ) 

"180 d\ 

for Dowson gas. 

For a transmission to a distance of 5,000 feet we get the 
following results : — 

POWBB Traksmitted IN Gas Maiks 
Length, 5,000 feet ; p^ >- terminal pressure = 14*7 lbs. per square inch ; 

initial velocity, 46 feet per sec 


Difference of 

Diameter of 

yni^in in jfia. 

Initial velocity 
in ft. per sec. 

Initial abGolnte 

lbs. per gq. in. 

pressTire pro- Quantity of 
duciog flow > gas in c. ft. 
in inches of i per hour 
water ' 

H.F. tnos- 

A, — Ordinury Lighting Oat 



45 1 15-60 24-7 31.815 

1,200 1 


— 1512 11-6 ' 127,260 




1490 5-5 509,040 




14-84 1 3-9 1,146,340 


3.—J)aw$on Gas 


45 I 16-71 55-5 




1 15-61 25-2 




— 1 1514 12-2 




— 14-99 8-3 1 1,145,340 

12,722 1 

Cost of Gas Engines and Dowson Gas Plant. — The following 
estimate * of the cost of a gas plant for an electric lighting 
station may be useful for comparison with the cost of steam 
plant previously given. 

Dowson gas plant for 1,160 iJi.p 


Seven 122 b.b.p. gas engines 


Two 61 bJi.p. gas engines 

Two 34 b,h.p. gas engines 


Dynamos and belting 


For 696 Kilowatts, total 


» Proo. Inst. C, -R, vol. cxii. p. 95. 




In 1877, Dr. William Siemens indicated the practicability and 
the probable commercial importance of the electrical transmission 
of power to considerable distances. In an address to the Iron 
and Steel Institute he stated that a copper rod 3 inches in 
diameter would transmit 1,000 h.p. thirty miles. In 1883, 
he delivered a lecture at the Institution of Civil Engineers on 
' The Electrical Transmission and Storage of Power,' but the 
only practically working transmissions which could then be 
described were the Lichterfelde and Portrush railways. About 
that time Marcel Deprez experimentally transmitted 3 h.p. a 
distance of 25 miles by ordinary telegraph wires, using a 
pressure of 2,000 volts, and obtaining at the motor only 32 per 
cent, of the energy expended. 

It has from that time been hoped that the transmission and 
storage of energy for motive-power purposes would be one of the 
largest fields of electrical enterprise. Much progress has been 
made, especially in the last five years. But, having regard only 
to plants actually at work, it must be confessed that the total 
amount of power transmitted electrically and used for industrial 
purposes, exclusive of traction, is not yet very great. In Mr. 
Gisbert Kapp's Cantor Lectures on the ' Electric Transmission of 
Power,' only one plant of considerable magnitude, that at Schaff- 
hausen, is described. At SchaflThausen two turbines, of 350 h.p. 
each, drive continuous current Oerlikon generators designed for 
an output of 330 amperes at 624 volts. Four cables, each of a 
section of 0*437 sq. in., convey the current 750 yards to actuate 
motors in a spinning mill. The power is sold at 21, IQs. per 
h.p. per annum. This installation is to be considerably ex- 
tended. In republishing his lectures Mr. Gisbert Kapp added 


a description of the Kriegstetten and Solothum installation for 
transmitting 50 h.p. at a pressure of 2,000 volts a distance of 
5 miles.* This is a continuous current system constructed by 
the Oerlikon Works. The commercial eflSciency, when 23 h.p. 
were received at Solothum, was 75 per cent. A number of other 
installations of a similar type, in which 50 to 300 h.p. have been 
transmitted 350 to 8,000 yards, have also been erected by the 
Oerlikon Company. 

A very interesting continuous current transmission was 
constructed in 1889 by M. Hillairet, of Paris, at Dom^ne, near 
Grenoble. A turbine of 300 h.p. drives a generator giving a 
current of 70 amperes at 2,850 volts. The current is conveyed 
5 kilometres to actuate a motor in a paper mill. The eflSciency 
of dynamo, line and motor is 65 per cent. The transmission 
has worked night and day with great regularity. 

In all these cases direct currents were used. In 1891 
alternating currents were employed in the striking LauflTen- 
Frankfort experiment. There 100 h.p., obtained from water 
power, was transmitted 108 miles with a loss of only 25 per cent. 
The experimental plant was erected by the co-operation of the 
AUgemeine Electricitats Gesellschaft of Berlin and the Oerlikon 
Company of Zurich. High tension ranging from 16,000 to 
30,000 volts was used in transmission. This was obtained by 
step-up transformers at Lauffen, and step-down transformers were 
used at Frankfort. The cost of the transmitting line has been 
stated at 15,000i. 

The special object of the experiment was to illustrate a 
solution of the problem of transmitting power by alternating 
currents. With the three-phase system, motors which are self- 
starting, without commutators, and very simple in construction 
can be used. The Lauffen dynamo also was of simple con- 
struction and mechanically of excellent design. At 150 
revolutions per minute it was capable of developing three 
alternating currents of 1,400 ampc^res each, at 50 volts above 
their common connection, equivalent to about 300 h.p. But it 
was usually worked at about 100 h.p. during the exhibition. 
The currents had 40 alternations per second. The conductors 
were of hard-drawn copper 4 mm. (0*16 inch) in diameter, having 

* See also report by Professor Weber, Die Leistun^ der Electriichen ArheiU- 
ahertra^ung ton Kriegstetten nach Solothum^ Zurich, 1888. 


a resistance of 2 ohms per mile. Each conductor served as a 
return to the other two. The line was carried on 3,227 wooden 
poles, the spans being 200 feet. The insulators were of porcelain, 
some of them having three oil grooves, but most only one oil 
groove. Both high and low tension circuits were grounded at 
the neutral point, or junction, of the three conductors. At 
Frankfort the current was transformed to 75 volts, and used 
pai-tly for incandescent lamps, partly to drive a 100 h.p. motor. 
'The greatest amount of energy transmitted is stated to have 
been 180 h.p. 

As an example of yet another system of transmission, the 
installation by the Westinghouse Company at the Gold King 
mine, Telluride, Colorado, may be mentioned. There a Pelton 
w^ater-wheel drives an alternating current generator. The 
^current is carried by an overhead line on posts a distance of 3 
miles, and actuates a synchronising motor of 100 h.p. The 
motor is started by a special Tesla motor. 

In these transmissions, and in nearly all hitherto carried out, 
one or more generators drive one or more motors belonging to 
a single industrial undertaking. In such cases much of the 
■difficulty and complication involved in a general distribution to 
many consumers are avoided, and the inconvenience and damage 
of a temporary stoppage, due to a breakdown of the line or 
lightning accident, are minimised. When the Niagara Com- 
mission met in London in 1891, only one case was known where 
power was distributed electrically to many consumers. That 
was the interesting installation at Oyonaz, not far fi*om Geneva, 
erected by Messrs. Cuenod Sautter & Co. Turbines of 250 h.p. 
at Charmines generate a continuous current at 1,800 volts, 
which is transmitted 8 kilometres by an overhead line to 
Oyonaz. There the current is reduced in pressure by motor 
transformers, and is distributed, partly for lighting, partly for 
driving small motors in a number of workshops. With a supply 
of cheap power the village was very prosperous, when the author 
saw it in 1892 ; but at that time only about 30 h.p. was 
distributed for power purposes, and 40 h.p. for lighting. 

In a great deal that has been said about the electrical distri- 
bution of power one thing has been too much overlooked. Means 
of transmitting power, even to considerable distances, have long 
been known. That they have not been more widely adopted is 


due, not to any risk of mechanical failure, but to the cost of 
transmission. If electrical transmission is to be extensively used, 
it must be when it can be carried out so cheaply, that power can 
be supplied at a less cost than that at which consumers can pro- 
duce it for themselves. 

Much has been accomplished in distributing electricity for 
lighting. But a higher price can be paid for electricity for 
lighting than for power purposes. Every Electric Lightings 
Company would be glad to supply current from its mains for 
power purposes, if only to increase the day load on the machinery 
and reduce the idle time. In Bradford, some electric motors are 
used for working hoists, lathes, &c. Recently, in London, electric 
motors supplied from the lighting mains have been applied to 
driving newspaper printing machinery. But the ordinary price 
of electricity for lighting purposes is 6d. per unit, which is 
equivalent to about 60Z. per h.p. per year of 3,000 hours. At 
that price it can only be used for power purposes, either when 
the power is required for short periods intermittently, or where 
there is great local inconvenience in employing steam or gaa 
engines. It is only where electricity costs from one-sixth to one- 
tenth of its ordinary price when used for lighting, that it can have 
any large importance as a means of obtaining power. 

In the application of electricity to traction on tramways and 
town railways, a remarkable success has been achieved. In the 
United States there are from 4,000 to 5,000 miles of electric 
tramways, for which the power is distributed from power 
stations. But for traction as for lighting a high price can be 
paid for power. 

For industrial purposes the question of cost is generally of 
controlling importance, and hence the progress of electrical 
methods of distribution has been less rapid. A review of the cases 
described above and others leads to the following conclusions 
as to the limitations of systems for distributing motive power 
electrically: — (a) When the power is initially steam power, its 
distribution electrically adds so much to its cost as to prohibit its 
transmission to any great distance in nearly all cases. (6) Hitherto 
it has only been in districts where cheap overhead conductors,, 
carrying high pressure currents, can be safely used that electrical 
methods of transmission have proved commercially successful. 

Conductors of Electricity. — The laws of electric flow are in 



many respects analogous to those of hydraulic flow. The resist- 
ance and the loss of pressure and energy depend on the length 
and section of the conductor. The amount of current which can 
be transmitted in any given case is limited by the heating effect, 
just as there are practical limitations to the velocity of flow in 
pipes. The permissible pressure in water mains depends on 
the strength of the pipes ; the permissible electric pressure on 
the insulation. 

A much wider range of electric pressureisallowable than can 
be permitted in fluid transmission. To carry current two miles at 
a given voltage, four times as much copper is required as to 
transmit it one mile, if the eflSciency of transmission is the same. 
Hence long-distance transmission would be enormously costly 
but for the possibility of varying the voltage. By doubling the 
electric pressure in the two-mile transmission, the amount of 
copper required is reduced to the same as for the one-mile trans- 
mission. Hence the whole problem of cheaply transmitting 
power to great distances depends on the use of high electric 

Comparison op Electbio Conductors 








of equal 



= 100 

= 100 

b I 

Tenacity, 1 

tons per Product of, 
square , a and 6 
inch I 

I : 


Pure copper 



Soft copper 

1 98 


Hard copper 



Swedish iron 

1 .16-5 


Galvanised iron 

' 14 


Cast steel . 

1 10-6 



' 55 


Silicon bronze 

1 97 


tt If 



it »» 

i *^ 


Phosphor bronze 

1 26 















22 5 



















Soft copper has the highest conductivity, and up to the 
present time it has been found to be the best material for con- 
ductors. Weight for weight, aluminium has nearly double the 
conductivity of copper, so that if its cost per lb. were less than 
half that of copper, which in time it may be, it would have 


almost equal merit. In the case of aerial lines, where in ectch 
span the conductor must sustain the stress due to its own 
weight,^ the tenacity is of importance. The comparative value 
of conductors for aerial lines is about proportional to the product 
of the tenacity and conductivity of equal weights. Here again 
aluminium is twice as good as soft copper and nearly as good as 
hard copper or silicon bronze. Perhaps some alloy of aluminium 
may be found not much heavier, of equal conductivity, and of 
greater tenacity than the pure metal. In that case the alloy 
might have an advantage over any material at present used. 

Laws of Steady Electrical Flow along Copper Conductors, — 
The weight of a copper conductor of a sq. in. section is, — 
w = 3 86 a lbs. per foot. 
= 20,380 a lbs. per mile. 

The electrical resistajice per unit length is, — 
r = pja 

where p is the specific resistance of the copper, which varies with 
its quality and temperature. For about 80° and copper of good 
quality, p = 0*0000086 if I is in feet, and p = 0*045 if I is 
in miles. A stranded conductor has about 28 per cent, more 
resistance, if a is understood to be the gross section, not 
deducting spaces between the strands. 

Let I be the intensity of the current or quantity of electricity 
circulating in amperes per second, in a conductor of length /, 
section a, with a potential difference E at the ends. Then 
E = r I I =z pil I a volts. 

The rate at which work is wasted in overcoming the resist- 
ance of the conductor is 

p i'^ I I a watts 
=^ p i^ 1 1 74t6 a horses power. 

When a generator drives a motor at a distance L there must 
be in general a going and return conductor, so that 2l must be 
substituted for I in the equations in finding the loss in the line. 

Case of Altematinfj Currents, — When an alternating current 
is transmitted the current is not uniformly distributed over the 

* Let w be the weight of the conductor in lbs. per foot run, I the span, 
and d the deflection in feet. Then the tension in the conductor due to its 
weight is — 

T = \%wPld lbs. per square inch. 


section of the conductor ; there is a tendency to accumulation 
towards the surface. Consequently for large conductors and 
high frequencies the resistance increases disproportionately to 
the weight of copper used. 

It is not possible to discuss the phenomenon here, but it is 
necessary to call attention to it as one of the special difficulties 
involved in the use of alternating currents. 

Heating of the Conductor, — The energy wasted in the con- 
ductor is expended in heating it, and the heat is dissipated by 
radiation and convection. The conductor takes a temperature 
at which there is a balance between the heating and cooling. 
The rise of temperature which can be permitted is limited by 
the increase of resistance, the decrease of insulation, and possible 
injury or danger if the rise of temperature is excessive. 

The most important experiments on the heating of con- 
ductors are those made by Mr. Kennelly in 1889. It is beyond 
the scope of the present treatise to consider these experiments in 
detail, and in the case of large conductors information is still 
defective. For cylindrical wires in still air the heating and 
cooling were found to be balanced when the following approxi- 
mate equation was satisfied : 

^^' = (0-073 fZ + 0-029) < 

where t is the rise of temperature of the conductor in degrees F. 
For blackened wires the cooling was twice as great. Professor 
G. Forbes has indicated that, for large conductors, thin strips 
will carry heavier currents than a solid conductor. 

Efficiency of a Generator Line and Motor, — Let E^ be the 
potential difference at the terminals of the generator ; E^ that 
at the terminals of the motor. Then, using the expression found 
for the line resistance, 

Ey — E^ = p I Z / a, 

which determines the section of the conductor when the loss of 
pressure in the line is fixed. 

If T is the electrical h.p. delivered by the generator, t,„ that 
received by the motor, 

T -T = PMi 

' " 746 a 


The efficiency of transmission is 

If rig 7),^ are put for the efficiencies of the generator and motor, 
taking account both of electrical and factional losses, the 
resultant efficiency of generator line and motor is 

Cost of Conductors, — In general, the method of transmission 
by aerial or underground, by bare or insulated conductors, will be 
decided by local or financial considerations. Then estimates can 
be made of the cost of conductors of assumed sections. Copper 
being an expensive material, a large part of the cost of the 
transmission (as distinguished from the generator and motor 
plant) is the cost of the copper and is proportional to the section 
of the conductors. There are some other charges which vary 
with the section of the conductors. But there are also charges 
which do not vary much with the size of the conductors. Hence 
the cost of the line of transmission per mile may be regarded as 
consisting of a constant part, and a part proportional to the 
section of the conductor. 

Mr. Stuart Russell * has given some estimates of the cost of 
conductors used in electric lighting. From these, simplified a 
little, some of the following formulas have been taken. Let a be 
the section of one conductor in square inches ; L the distance of 
transmission in miles, so that for a going and return conductor 
the length of the conductors is 2 L : then the cost in pounds 
per mile of the line of transmission is: — . 

Cost per mile in 

Insulated cable in iron pipes or bitumen casing . 897 + 4,1 ISa 

Armoured cables in the ground .... 7l>2 + 4,43oa 

Concrete culvert with bare conductors . . . 1,584 + l,7D5a 

Bare conductors on iron posts, with insulators for 

lines of large capacity 500 + 1,000a 

Bare conductors on wooden posts, with insulators 

for light lines 20 + l.-lOOa 

These formulao give the cost of the line erected, but 
difierences of cost of copper, labour and carriage may involve 
considerable differences of cost in particular cases. 

> Electric Light Cablet. London, 1892. 


Condition determining the most Economical Section of Con- 
ductors. — Suppose current has to be conveyed from a generating 
station to a point of consumption at a distance L feet. In general 
the system to be adopted will be predetermined, and the cost 
of generating the current can be ascertained. Suppose, further, 
the kind of transmissson selected so that the cost can be 
expressed in terms of the section of the conductor. If the fall 
of potential in the line is fixed, then the equation above gives 
the section of the conductor. The minimum section of con- 
ductor is fixed with reference to the heating permitted. Subject 
to this limitation, in other cases, financial considerations govern 
the size of the conductor. In a temporary installation the total 
capital cost of generator and line should be as small as possible. 
In some cases of generating by water power the total amount 
of current which can be generated is fixed, and then 
considerations as to the way in which the energy is to be dis- 
posed of may impose conditions on the amount of waste in the 
conductor. But most generally, in cases of extensive power trans- 
missions, an indefinite amount of power is available at a fixed 
cost per h.p. at the generating station, and as much of it as is 
not wasted in the conductor can be disposed of at the end of 
the transmission, at a fixed charge per h.p. The consideration 
which then determines the most economical section of the con- 
ductor is this : the larger the conductor, the less waste of 
energy will occur in transmission, and the less will be the cost 
of producing that energy. On the other hand, the larger the 
conductor the greater will be that part of the cost of the trans- 
mission which depends on the section of the conductor, and the 
greater the annual charge on that part of the capital expended. 
Lord Kelvin pointed out, in 1881, that in such a case the most 
economical section of conductor is that for which the annual 
cost of the energy wasted in transmission is equal to the annual 
interest and depreciation on that part of the cost of the trans- 
mission which is proportional to the section of the conductors. 

Case in which the Current is of Constant Intefisity in WorMng 
Hours, — For simplicity suppose one generator driving one motor 
L miles distant, the total section of the conductors being 2a, Let 
the capital cost of the line of conductors be expressed in the form 

(a 4- ^ d)^ pounds 


where a and /8 are constants, some values of which are given 
above. If s is the rate of interest and depreciation charged on 
the capital expended in pounds per pound, 

€ (a + j8 a) L pounds 

is the annual cost of the line of transmission. Now let g be the 
cost of producing one h.p. hour of electrical energy at the 
generating station, including interest on plant, maintenance, and 
working expenses ; T the number of hours per annum during 
which energy is supplied. The cost g will be greater as the 
number of hours worked is less. From the equations above, the 
energy wasted in transmission will be 


- i- h.p. 

746 a ^ 

The annual cost of the energy wasted will be 

2 pi'^hgi 
" 746 a " 

The total annual expenditure in transmission will be 

/ . o \t . 2pI^L7T 
/ lu a 
which is a minimum, if — 

/40 a- 

1= /f373^f 

which gives the amptires per sq. in. of section of conductor. If 
the section is determined in accordance with this law, the annual 
cost of energy wasted is equal to the annual charge for interest 
and depreciation, on that part of the cost of the line of trans- 
mission which is proportional to the section of the conductor. 
The adoption of this density of current is subject to the con- 
dition that it does not cause excessive heating of the conductor. 
The current density is independent of the length of the trans- 

Case of a Varying Current, — The energy wasted in a given 
conductor is proportional to the square of the current. Suppose 
that in a given transmission the maximum intensity of the 


current is i, and that the current x^ i is transmitted for t^ hours ; 
ajj I for <2 hours; and so on. Then the equivalent constant 
current, or constant current for which the waste of energy would 
be the same as that of the actual varying current, is 

/ UH,-^xX+ ) 

where T = ^, + ^^ + . . . the whole yearly hours of transmission. 
It is this equivalent current which is to be used in applying the 
equations above. 

In the following calculations i is to be taken as the current 
intensity in each conductor, if the current is constant, or as the 
equivalent current intensity for the actual working hours, if the 
current varies. 

Calculation of Proportions and Cost of Electrical Transmissions. 
In order to see how the cost of electrical transmission varies 
with local conditions of power available and distance, and with 
electrical conditions of voltage and system of transmission, a 
series of cases has been calculated. A simple transmission be- 
tween generator and motor without complex distribution is 
assumed. The interest on cost of dynamo is supposed to be 
included in the cost of the power. That on the cost of motors 
is not taken into account, nor that on- the cost of transformers, 
if required. 

First of all, it is obvious that, speaking broadly, only those 
cases are suitable for electrical transmission in which the 
economical conditions governing the size of conductor can be 
complied with. In the following table the current density 
(amperes per sq. in. of conductor) has been calculated for 
various values of a h.p. year at the generating station, and 
for three assumed rates of interest on the capital cost of the 
line of transmission. The value of a h.p. year has been taken 
at 0-5Z. to lOZ. Taking the working year at 3,000 hours, this 
would correspond to a value per h.p. hour, for the power 
generated, ranging from jr=004tZ. to r/=0'8(Z. The cost of a 
mile of two conductors, excluding that part which does not vary 
with the Section of the conductors, has been taken at )8=1,795Z. 
when the two conductors have two sq. ins. of section. 



Table I. — Cukkent Density 
» = 1J95/. 

1 ^T = 0O8tof ll.p. 

' year at generating 
sUtiou in Um. 

i/a = carreuc uen 


*iry m amperes per square i 

Qca, lor rat 






1,050 1 






1 40 ; 


525 1 










1 100 




The largest of these densities would probably involve ex- 
cessive heating of the conductor. With these exceptions the 
numbers are practically suitable. 

The following table gives the number of ampdres in the 

circuit for different amounts of h.p. transmitted from the 

generating station, at various potential differences E^ at the 
terminals of the generator. 

Table II.— Intb>'8ity of Cubbext in Cibcuit 
H.P. = Ey 1/746. 


trail smitteii 


1 100 

Current i in i 



unpdre* for pressure in volts 
2,500 1 5,000 




30 ' 







75 ' 




, 1.492 


149 ^ 









' 74,600 





For some of the cases given in this table, the following are 
the sections in square inches of eacli conductor of a line of trans- 
mission : — 



Table III.— Section op Conductors 

H.P. transmitted 
from generator 

Economical section a o/ one conductor in sq. Ins. for pressures 
at generator lu voltt) 




Case 7.— Cost of line of transmLssion proportional to section of 
conductor =: /3 == 1,600/. per mile; interest on cost of line 
7J per cent, per annum (« « -075) ; cost of h.p. year at 
generating station = yT = Ool. 








1-93 -386 


• 039 



3-86 -772 





lJ»-3() 3-86 





38G2 772 




Case II.— Aa above, but cost of I 

i.p. year « , 

fi - 21. at gene- 

rating station. 


•772 154 


-01.5 •OOS 


3-86 -772 





7-72 ' 1-64 





38-60 7-72 





77-20 15-44 




Case III.— As above, but cost of 1 

ti.p. year -^ 

7T = 6/. at gene- 

rating station. 































The loss of pressure in the line in volts can now be calcu- 
lated for the cases given in the last table. 


which, for a mile of transmission and for the value of i/a given 
by Lord Kelvin's law, becomes 


L )«!( of pressure 
lu vults per mUe 

Case! 127 

„ II 63-5 

„ III 36-6 

T 2 



Suppose it is assumed that, when more than 25 per cent, of 
the pressure at the generators is wasted in consequence of the 
resistance of the line, the conditions of transmission are unsuit- 
able. Then some of the cases in the table above will be of this 
class. The conditions will be unsuitable if the voltage at the 
generator is less than 

Case I. 
M II. 

n HI. 

5 uiila 30 mile* 

2,540 10,100 

1,270 5.0SO 

732 2,9>8 

Subject to this limitation, the amount of h.p. delivered at 
the motors can be calculated for the cases in the table above. 

Table IV. — H.P. Delivebed at Motobs 

! Five-mtle transmission. 
H.P. at 1 Volts at generator 




2,600 6,000 

























Twenty-mile transmission. 
Volts at generator 



2,500 6,000 , 10,000 














Case II. 
















97; — 



76: 87 

484 i — 


373 1 436 

968 — 



746 1 873 

4,841 — 



3,730 43,65 

9,683 — 


1 — 

7.460 8,730 






500 — 




1,000 — 














Case III. 

98| - 

491 ' - 

982 1 - 

4,909 - 

9,817, - 

I 427! 

4,272 I 

93 I 





The following table gives the cost of line per mile of transmis- 
sion in the cases already considered. For transmissions of 100 to 
1,000 h.p. at the generating station, the cost is calculated by 
the formula, cost = 20 + l,600ci pounds per mile, the conductors 
being supposed to be carried on wooden posts. For trans- 



missions of 5,000 and 10,000 h.p., cost = 500 + 1,600a, the 
conductors being on iron posts. The distance of transmission 
does not affect the cost per mile. 

Table V. — Cost op Line per Mile op Transmission 

CoBt in pounds per mile for pressures in volts at the 

H.F. ftt 

generators of 


2,500 1 6,000 1 10,000 






















Case IL 



44 1 33 1 















2,970 1,735 

Case III, 



















4,782 2,641 

Finally, the cost of a h.p. year delivered at the motors can 
be calculated. The interest on the cost of the line of trans- 
mission is taken at 7^ per cent. ; and the cost of a h.p. year at 
the generators is that given in the statement above, namely, 
Casel.,r/T=0'5 ; Case II.,^T=2 ; and Case III., ^ T= 6 pounds 
per year. 

The following table is instructive, even if every allowance is 
made for the extent to which local circumstances, cost of carriage, 
and difficulties of various kinds in construction, may affect the cost 
per mile of a transmission. In the first place, it is clear how 
important high electrical pressure is in making economy of cost 
possible in long transmissions. Next, it is useful to note that, 
when high pressure can be adopted, the transmission adds so 
little to the cost of the power delivered that a greater expendi- 
ture is quite justifiable, if it secures more safety and security from 



accident. In high pressure transmissions greater expenditore 
on the line of transmission would not add so much to the cost 
of a h.p. at the motors that the use of the power would be 
seriously restricted. If high pressui-es are used, the simplest 
precaution to ensure safety against accident to life is to fence 
and patrol the line of transmission. This involves purchase of 
land and other expenses not included in the estimates given 
in Table VI. 

Table VI.— Cost of a H.P. Year at Motors 

H.p. at 


Five-mile traiidmiijuion 


Twenty-mile tranamifidon 

Cost ia pounds of a li.p. year at motors for pressures 
iu volts lit generators of 





Caie I, — Cost of a h.p. year at geDeratore 0-5i. 































Case J I 

—Cost of 1 

ft h.p. year 

at generators 









2 52 







2 24 








3-38 , 






3-28 1 

2 59 

Case II. 

r.— Cost of 

a h p. yea 

r at generators 60^. 



1 6-49 


8-17 • 









1 G-83 

i 6-40 







6 23 

, 7-95 






7 86 


I Wood 
I posts 

1 Iron 
/ posts 

'1 Wood 
I postA 

1 Iron 
I j posts 

I posts 

\ Iron 
J posts 

System of Electric TransviUting Maitnt. — For continuous 
currents the following arrangements are possible : — (a) Single 
conductor and earth return; (b) Going and return conductor 
of equal size; (c) Three-wire system, one being a balancing 
wire. The earth return is objectionable, partly because it 
interferes with telephone and telegraph systems, partly because 
it is dangerous to life, if the potential exceeds 500 volts. 
Method (h) is most suitable where there is one generator and 
one or more motors not distant from each other, and for distribu- 
tion in series. In other cases method (c) has advantages. It 


would seem best to have the three conductors of equal size, 
so that, in case of accident to one conductor, the two others 
could be used to supply current, as in method (6). At the 
station where the three-wire transmission terminates, the distri- 
bution to local circuits can be from time to time re-arranged 
so as to keep the current in the balancing conductor small. 

For alternating currents of single phase, a going and return 
conductor may be used. For alternating currents of two phases 
it is possible to transmit by three conductors, but it appears 
preferable to use four conductors of equal size. The three- 
phase system used in the Frankfort-Lauffen experiment requires 
three conductors of equal size, either acting as a return to the 
two others. 

In all distributions for power purposes to a distance hitherto 
carried out, except some mining installations, bare conductors 
carried on wood posts have been used. In cases like Oyonaz, 
where the line is carried over fields to a small village, such 
a cheap method may be used without much objection, even 
when high pressure currents are transmitted. In that case, 
however, there is a liability to injury, especially to injury from 
frost and sleet damaging the insulation of the line, and to 
injury from lightning, which must be reckoned with. Malicious 
damage is also possible with aerial transmissions. , In important 
aerial transmissions a patrol to detect and remedy defects seems 
necessary, and adds to the working cost. There are difficulties 
in using insulated cables for high pressures, besides their cost 
and tlie liability of the insulation to deteriorate. 

It is fair to point out that, when the cheapness of electrical 
transmission is put forward as a reason for adopting it in 
preference to other methods of transmitting power, it is always 
assumed that such a rough expedient as overhead conductors 
on wooden posts can be adopted. To a mechanical engineer 
such arrangements do not appear to afford adequate security 
or permanence for an important power distribution. In pro- 
portion as the number of consumers taking power from a common 
source becomes greater, the inconvenience, cost, and damage 
of any temporary stoppage of the supply of power become 
more serious. Hence it will probably prove to be necessary, 
if any general system of distribution of power by electricity 
is carried out, to place the conductors in subways, where they 



are protected from injury. Such a constniction, however, will 
necessarily increase the cost of electrical transmission. 

The smallest self-respecting town requiring a water supply 
would not hesitate to build such a concrete conduit as that 
shown in fig. 74. D'Arcy built such a conduit 13 kilometres 
in length for the water supply to Dijon. It is the smallest 
conduit accessible throughout. An important electrical power 
distribution needs permanent and secure construction as much 
as a system of water supply. An objection is sometimes made 
to a subway for bare conductors carrying high pressure currents, 

that there would be danger 

' ^m^^f^^^^f^ff^^^^^yj^hfi^MjiJfiJi'* to life in ti-aversing the 

"■ " ^ • '* * ^ ^ ^ "^ ^r"5- •- conduit. To obviate danger 

as much as possible the con- 
ductors have been placed 
in recesses. Further, by 
movable metal screens 
put in connection with a 
return or earthed conduc- 
tor any part of the conduit 
could be made absolutely 
safe while repairs were in 
progress. The figure shows 
only a rough sketch of a 
possible arrangement, but 
some permanent protection 
for conductors will have to 
be adopted in important 
electrical distributions. 
Electrical Systetm. — The continuous current method of 
electrical working has hitherto been most frequently adopted. 
The alternating current produced in the dynamo is commutated 
into a continuous current for transmission, and commutated back 
into alternating currents in the motor. Direct current dynamos 
and motors are well understood. The motors are satisfactory, 
start with a load on, and have been largely used for tramway 
and other purposes. For transmission to moderate distances 
the direct current method is well adapted. On the other hand, 
it has two essential defects or limitations. The first is that the 
electric pressure of the direct current can only be altered by 

Fig. 74. 


expensive running machines termed motor transformers, or 
motor and generator combined. If, for transmission economically,, 
a high pressure current is used, it can only be reduced in 
pressure, for distribution in places where the high pressure is 
dangerous, by motor transformers. Next, there appears to be a 
limit to the electrical pressure which can be obtained in a 
direct current system, due to the complicated construction and 
diflSculty of insulation of the commutators. It does not seem 
practically possible to obtain more than 2,000 or 2,500 volts ^ 
in direct current dynamos. By coupling these in series a higher 
pressure can be obtained, but this involves new insulation dif- 
ficulties, and probably 10,000 volts is the limit at which a con- 
tinuous current system is likely to be worked. 

There are two arrangements of direct current systems. 
Ordinarily the current is transmitted from the generatora at 
constant pressure, and the motors are connected with the mains 
in parallel. Motors can be connected to the mains or discon- 
nected without interference with others, unless the line resist- 
ance is excessive. Working in parallel the pressure in the 
mains is nearly constant, and the loss of energy in the line 
diminishes as the load diminishes. 

The second system of direct current working is to transmit 
a constant current, with variable electric pressure, increasing as 
the load increases, the motors being all in series.. If all the 
motors are disconnected, one dynamo is run at a voltage just 
sufficient to send the full current through the mains. As motors 
are added and the resistance increases, the dynamo is run faster 
and the voltage increased. When one dynamo is working at 
full speed, a second is run on closed circuit till it is delivering 
the required current. It is then connected in series with the 
dynamo already coupled to the mains. If the load still increases 
the speed of the second dynamo is increased, and when it has 
reached full speed a third dynamo may be added in the same 
way. With this system the loss of energy in the mains is the 
same at all loads. It is not, therefore, so efficient as the parallel 
system, but it has in other respects advantages. The regulation 
of the dynamos, so as to maintain a constant current by varying 
their speed, is comparatively easy, and the regulation of the 

' l^essures as high as 4,600 volts have been proposed in direct current 



motors to constant speed by a centrifugal governor regulating 
the excitation is also easy. For small motors the parallel 
system is more convenient. 

Messrs. Siemens Brothers & Co. proposed a series system of 
this kind for the Niagara distribution in 1890, and Messrs. 






Fig. 76. 

Cuenod Sautter & Co. have since carried out a very important 
power distribution at Genoa in the same way. 

Fig. 75 shows an ordinary constant pressure, continuons 
current distribution with motors in parallel. There is an 
economy in copper, however, if a three- wire system is adopted. 
For a long distance transmission it would appear best to have 
the three conductors of equal section, so that any two could be 


Fig. 76. 

used to transmit current in case of accident to one. Fig. 76 
shows the constant current system of distribution with motors 
in series. 

During the last three years the alternating current method 
of electrical working for power purposes has made considerable 
progress. In this method the alternating currents produced in 



the dynamo are transmitted without commutation. The insula- 
tion difficulty which limits the pressure in direct current 
dynamos is much less serious. Further, the electric pressure 
of an alternating current can be altered to a higher or lower 
pressure, with great facility, in inductive transformers having no 
moving parts, requiring no attention and easily insulated for 
almost any pressure. This facility of varying the electric 
pressure for different purposes is of enormous importance in a 
general system of electric distribution. 

Up to a recent period alternating motors did not meet all 
requirements. An alternating dynamo generator will run as a 
motor, if supplied with current, synchronously with the generator 
supplying the current. But it has no starting torque, and 
requires to be put independently into motion at the right speed 



^^ MOTOn 

Fig. 77. 

before current is supplied to it. Once started in synchronism, 
it will keep step with the generator, if of suitable type, even if 
•considerably over or under loaded. But the necessity of starting 
synchronous motors by an independent motor is for many pur- 
poses a serious defect. Next, the rotary field motors of Tesla 
And Ferraris were developed. For these two or more alternating 
-currents differing in phase are required. These motors start 
with a load, and are now made both of a rotary field non- 
synchronous or of a synchronous type. More recently still 
motors have been constructed, which are self-starting, with an 
ordinary single-phase alternating cuiTent. 

Fig. 77 shows the arrangement of an alternate current two- 
wire, or single-phase, transmission. The principal motor is a 
fiynchronising motor. But as this is not self-starting, a Tesla 



single-pbase motor is used as a starting motor. The startiDg^ 
motor is first put into circuit. When this comes up to speed, it 
is used to drive the synchronising motor by a friction clutch. 
When the synchronising motor is at precisely the right speedy 
which is indicated by a special instrument (termed the synchro- 
niser), the starting motor is disconnected, and the synchronising 
motor put into circuit. The operation can be carried out in two- 
or three minutes. 

The Westinghouse Company in the United States use in 
ordinary cases for transmissions of this kind a pressure of 3,000 
volts and 1 20 alternations per second. The commercial eflSciency 
of generators and motors varies from 88 per cent, in small sizea 
to 92 per cent, in large sizes. That is, a combined generator 







«TC» UP 

Fro. 78. 

and motor would have an efficiency of 77 to 81 per cent., ex- 
clusive of loss in transmission. These efficiencies are at full 
load. At quarter load they are about 30 per cent. less. 

Fig. 78 shows the arrangement of an alternate current 
polyphase system. Each generator delivers two currents to 
step-up transformers. These currents may be of low pressure, 
so as to be handled safely. After transmission, at a sub-station, 
the current is lowered in pressure by step-down transformers. 
The motors shown are two-phase Tesla moturs which are self- 
starting. But a motor generator' is also shown, producing a low 
pressure direct current for working a tramway. Three con- 
ductors could be used to each generator, but it appears to be 
preferable in long distance transmission to use four. Further, it 



is now certain that alternate current dynamos can be so con- 
structed that they can be worked in parallel, even when driven 
by separate turbines or engines. In that case two or more 
generators may deliver current into the same four conductors ; 
the dynamos once being run to the speed at which they fall into 
step will not break step either with large variations of the 
driving efforts, or the excitation, of the different generators. 
With a two-phase system currents for lighting or for working 
single-phase motors may be taken from a pair of conductors. 



iTCP DOWN rnAHirowcn 





Fig. 79. 



i/00-«00 VOLTS 

Fig. 79 shows the arrangement of the three-phase alternate 
current system in use at Ileilbronn. The * Drehstrom ' dynamo 
produces current at 50 volts, which is transformed for trans- 
mission to 5,000 volts. This is transformed down in Heilbronn 
in two stages to 1.500 and 100 volts. In the village of Sontheim 
it is transformed directly to 100 volts. There are three equal 
conductors carrying current in different phases, and all con- 
nected at a common neutral point. 





In general it will not pay to transmit power by electrical 
methods when the energy has to be generated by steam power. 
The exceptions are when special economies or conveniences- 
result from electrical transmission, or where a high price can be 
paid for power, as in the case of tramways and town railways. 
One case where electrical methods have proved to possess some 
convenience is in distributing power to different workshops not 
widely distant from each other. The prime motor generates 
current which actuates secondary motors, conveniently located in 
the workshops, and thus cumbrous mechanical transmission is 
avoided. Generally, however, it is where water power is avail- 
able as a source of energy that electric transmission of power 
can be most advantageously applied. The progress accom- 
plished in this direction will be best explained by describings 
some typical cases. 

It will be found that, for some time, direct current methods 
were alone used. But the need of higher electric pressures than 
are possible with the direct current system has compelled 
electrical engineers to study the construction of motors suitable 
for alternate currents. The synchronous motor has been well 
understood for some years, but there are obvious objections to 
the general use of a motor which can only be started by some 
independent motor. The three-phase motor of Brown and 
Dobrowolsky was the first alternate current motor which seemed 
to possess the practical advantages of continuous current motors. 
These advantages, however, are gained at the expense of com- 
plication in the line. The two-phase motors of Tesia have now 
been constructed in large sizes, and appear to meet most of the 


conditions of practical application. In small sizes, at any rate, 
single-phase, non-synchronous, alternate current motors have now, 
been constructed, by Brown, Tesla, and others, so that the initial 
difficulty of using alternate currents, that there were no trust- 
worthy motors suitable for ordinary purposes, seems likely to 
disappear. The multiphase motors have the advantage over 
direct current motors that they have no commutators or brushes^ 
and are simpler in design, while they have in common with 
them the ability to start against a load and to run non-synchron- 
ously, which is often an advantage. As to the present position 
of the question of the choice of an electrical system, Mr. A. Sie- 
mens said, in his recent address to the Institution of Electrical 
Engineers, that * no hard and fast rule can be laid down that 
either the direct or alternate current system should be preferred 
in all cases, where power has to be transmitted to a distance. 
In judging of the merits of the two systems the proper con- 
clusion can only be arrived at if the commercial aspect of the 
case is allowed to decide the question.' 


Tlie Herstal Smull Arms Factorij hi Belgium, — In 18G6, the 
manufacturers of small arms in Belgium fonned a syndicate, 
and, in order to carry out a large Government order, decided on 
the erection of new workshops of the most modem construction. 
M. Leon Castermans has given an account ^ of the development 
of the plans for these works, and of the reasons for the adoption 
of electrical transmission. 

The operations to be carried on involved the construction of 
a number of different factories, so arranged as to be capable 
of future extension. In these factories it was found necessary 
to provide for 13 lines of shafting requiring a total of 200 h.p. 
or, allowing for loss in transmission and engine friction, about 
300 i.h.p. Electric lighting was next decided on, and for this 
an additional 160 i.h.p. was needed. 

It now became a question how these lines of shafting were 
to be driven from the steam engine. The mechanism between 
the steam engine and the lines of shafting in the workshops 

' Jfevue VniteneUe da Mines, xix. 1892. There is an account of the 
Heretal Factory in the Engineer, November 26, 1892. 


may conveniently be termed tlie intermediate transmission. 
This is required to sub-divide and distribute the motive power, 
And to eSect necessary modifications of speed. The whole 
amount of power to be transmitted was not very great, and it 
had to be much sub-divided. If ordinary mechanism were used 
there would be a loss of power at each step of the process of 
■distribution. The aggregate of these losses is large. It was a 
special inconvenience of a mechanical system of distribution 
that no part of the mechanism of transmission could be dis- 
engaged, except by the use of somewhat cumbrous appliances. 
Hence, practically, the whole transmission would be kept 
running even when part of the workshops were idle. The waste, 
considerable at full load, would be largely increased when only 
part of the machines were at work. 

Two systems of transmission were first studied — shafting and 
gearing, and rope transmission. It was found that, for either 
system, there would be required some 30 tons of pedestals and 
fixed supports and 40 tons of moving mechanism. Practically, 
the whole of the moving pieces would be constantly running, 
whether much or little work was being done. There is a further 
disadvantage of such systems — they do not easily lend them- 
selves to extensions of the works. Either the gearing must be 
initially of excessive size, or when an extension is necessary it 
must be removed and replaced by new gearing. 

These considerations led to an investigation whether electric 
transmission could be adopted, with secondary motors driving 
each line of shafting. 

Finally, it was decided to have one dynamo of 500 h.p., with 
its armature directly connected to the crank shaft of a steam 
engine ; to transmit the power electrically, and to have secondary 
electric motors for the lines of shafting and some special tools. 

The mechanical efiiciency of the engine has been found to be 
94 per cent. The dynamo has an eEBciency of 90 per cent., and 
the motors have an average efiiciency of 87 per cent, at full load. 
Allowing two per cent, loss in leads, the resultant efficiency is 
0-94 X 0-87 X 0-90 x 098 = 0-72 ; or 72 per cent, of the 
indicated power is delivered to the lines of shafting. 

The electric system has the advantages — (1) simplicity in 
the transmission between the dynamo and secondary motors ; 
»(2) saving of waste of work in consequence of the readiness with 


which any secondary motor can be disconnected; (3) great 
eflSciency of transmission ; (4) facility for future extensions with 
little modification of existing plant. 

The installation has been very successfully carried out. The 
steam engine of 500 h.p. was built by the Society Anonyme, 
Van den Kerchove, of Ghent. It runs at 66 revs., and carries 
the armature of the dynamo on its crank shaft. The armature 
serves as a fly-wheel. The field-magnets are shunt wound. 
The dynamo can develop 2,400 amperes at 125 volts. There 
are two commutators, each taking half the current. 

There is no doubt that, in factory driving, the distribution of 
secondary motors economises power. But it is fair to point out 
that the problem of distributing power by secondary motors in 
fiuch cases did not first arise at Herstal, and can be solved 
with somewhat similar advantages by other methods. It has 
alwsfys been a question in large works, how far it is desirable to 
have a single steam engine, or smaller engines driving special 
parts of the works. In dockyards there has long been distribu- 
tion of power to secondary motors by pressure water, or com- 
pressed air. In some works scattered gas engines have replaced 
a single steam engine, and much of the intermediate trans- 
mission is then dispensed with. For a long time at Seraing, 
and recently at some works in America, there has been a 
distribution of power by compressed air to secondary motors 
working special departments or machines. 

Transmission for Mining Work in Nevada, — At the Gomstock 
mines there existed a Pelton wheel of 200 h.p., on a fall of 
460 feet. To obtain additional power, the water discharged 
from this wheel is conducted by two iron pipes down the 
vertical shaft and incline of the Chollar Mine, to the Sutro 
Tunnel level. It is there delivered under a head of 1,630 feet to 
«ix 40-inch Pelton wheels. Each wheel develops 125 h.p. at 
900 revolutions per minute, the jet of water being only f of an 
inch in diameter. Each Pelton wheel is coupled to a Brush 
direct current dynamo. From these generators the current is 
taken to the surface, where it drives motors which work a 60- 
stamp mill. The eflSciency of the electric apparatus is said to 
be 60 per cent., and the Pelton wheels weigh less than 2 lbs. per 
-electric h.p. 

Traiisraissiori at the Oreenside Silver-lead Miiu^s, Vatterdale^ 



Wesi/moreland. — ^Water is supplied from a small tarn under 
Helvellyn through 15-inch pipes, and works a 100 h.p. vortex 
turbine under a head of 400 feet. This drives a four-pole com- 
pound direct current dynamo. The current at 600 volts is 
taken 6 furlongs by bare conductors on poles, and thence into 
the mine by insulated cables. In the mine there is a 9-h.p. 
series motor for winding, a 9-h.p. motor for pumping, and a 
motor transformer giving a current at 250 volts, which works 
motors on a tramway. There is also some lighting by incan- 
descent lamps in aeries of six. 

The Genoa JmfUdUitiou. — Exceptional circumstances have 
made it possible, at Genoa, to establish an electric supply in con- 
nection with a water supply. Some ten or twelve years ago, 
works for supplying water to Genoa additional to others then 
in operation were constructed. The water is obtained from 
streams on the Piedmont-Liguria frontier and impounded in 
reservoirs on the Gorzente River, an affluent of the Po. The 
reservoirs are 2,050 feet above Genoa, there being a large 
surplus fall not required as head for the water supply. To 
relieve the pipes of unnecessary pressure three relieving or 
service reservoirs were constructed, reducing to about 600 
feet the pressure available for the transmission of the water 
to Genoa. The firet reservoir is near the outlet of a tunnel by 
which the water is carried through a mountain ridge, and is 
360 feet below the tunnel. At that point 730 gross h.p. 
can be utilised. The second reservoir is 360 feet below the 
firat, and there also 730 gross h.p. can be obtained. The 
third reservoir is 500 feet below the second, and there 1,000 
gross h.p. can be obtained. The water supply scheme has 
not been entirely successful, and the engineer, M. Bruno, and 
the consulting engineer, M. Prdve, were led to consider the 
utilisation of the surplus fall to generate electricity to be 
transmitted for lighting and power purposes to Genoa, distant 
about sixteen miles. 

At the three points described electric generating stations, 
named after the Italian electricians Galvani, Volta, and Pac- 
cinotti, are now in operation. 

A first installation of a turbine of 140 h.p. was made in 
1889 at the Galvani Station. This proved successful, and the 
further development of the stations was undertaken. The 


Galvani Station supplies electricity to fifteen motors distri- 
buted along the valley from Iso^erde to Genoa, and to a 
motor of 60 h.p. at the railway station in Genoa. It also 
supplies motor transformers at the Central Electric Lighting 
Station in Genoa. The remainder of the power at this station, 
amounting to 600 h.p., is utilised by means of telodynam.ic 
transmission. The Volta Station supplies electricity for 
lighting the station of Sempierdarena, by a motor of 60 h.p. 
driving twelve Siemens and two Technomasio dynamos, and 
also supplies electricity for motive power to a number of mills 
and factories, and the repairing shops of the railway. More 
recently the Paccinotti Station has been put in operation, and 
there are now two circuits, one fed from the Volta, the other 
from the Paccinotti Station. Messrs. Cuenod Sautter & Co., 
of Geneva, who installed the electrical plant, have furnished 
the following details. Messrs. Rieter Brothers, of Winterthur, 
constructed the first turbines erected. The remainder have 
been constructed by Messrs. Faesch & Piccard, of Geneva. 

Electrical System. — The distribution is in series at constant 
current. The generating dynamos maintain a current of con- 
stant intensity in a single circuit which traverses all the motors. 
They supply a constant number of amp6res, whatever the number 
of motors at work. The voltage is essentially variable. At cer- 
tain hours all the motors are out of action ; then the dynamos fur- 
nish a current at 450 or 500 volts, corresponding to the loss in 
the circuit. At certain hours both the motors supplying power 
and the motors driving dynamos for lighting are in action ; then 
the voltage reaches 5,000 to 6,000 volts. 

Oalvani Station, — ^This has a smgle group of machines, . 
consisting of two Thury continuous current dynamos connected 
by RafFard couplings to a Rieter turbine of 140 h.p., having a 
normal speed of 450 revs, per minute. In addition, a jate 
factory absorbs the power of two Rieter turbines of 300 h.p., 
the power being transmitted to it by telodynamic cables. The 
generators have six poles and give at full load 47 amperes at 
1,000 to 1,100 volts. Their speed varies from 20 to 475 revs. 
per minute, according to the demand for power. The dynamos 
are coupled in series and work day and night. 

Volta Station, — This has been in operation since 1891. 
There are four turbines of 140 h.p. each, and eight dynamos 


working at 47 amperes and 1,000 volts at full load. These 
generators work at constant speed, and the regulation is effected 
by varying the exciting current. The regulation is effected 
by a single regulator, however many generators are in action. 
The main turbines are Faesch & Piccard turbines, with relay 
governors which maintain a strictly constant speed. 

The regulation of the exciting current involves difficulties, 
because the voltage of the main circuit varies from moment to 
moment, and sometimes quite suddenly, when motors driven by 
the current are thrown out of action. The motors are thrown 
out instantly by short-circuiting them. To meet these conditions 
the exciting dynamo is driven by a separate 15 h.p. turbine, 
which has as little inertia as possible. The exciting dynamo has 
a very light armature, so that it follows instantly the variations 
of speed of the turbine driving it. The exciting dynamo is 
itself excited by a small machine serving to light the station, and 
the stability of its magnetic field is thus independently secured. 
The turbine driving the exciting machine is provided with a re- 
lay governor, but the conical pendulum ordinarily used to secure 
constant speed is replaced by a solenoid, holding in equilibrium 
a soft iron core weighing 33 lbs., which is directly attached to the 
valve of the relay. A leather belt keeps this core and the valve 
in rotation, so as to practically annul any frictional effect. A 
spring and counterweight permit the adjustment of the action 
of the regulator. The solenoid is traversed by the current in 
the main circuit, which is normally 47 amperes. If the current- 
augments, the core of the solenoid rises, puts in action the 
relay, and closes the sluices of the turbine driving the exciter. 
Vice versd, if the current decreases the relay opens the turbine 

Tlie Paccinotti Station. — The system at the Volta Station 
has given good results, but at the Paccinotti Station a return 
has been made to a system similar to that at the Galvani 
Station. This station has four groups, each consisting of one 
turbine of 140 h.p. and two dynamos. The turbines have no 
speed governors or fly-wheels. The dynamos are self-exciting, 
with very light moving parts. The regulation of the current 
is effected thus. The turbine sluices are actuated by Piccard 
relay motors. The slide valves of the four relay motors can 
be separately connected to or disconnected from a shaft running 


the whole length of the building. This shaft is driven by a 
1 h.p. electric motor, which has on its armature two windings 
in opposite directions. Each winding has its own commu- 
tator. The problem is then to act on the motor so that it shall 
revolve in a direction causing a closing of the turbine sluices 
when the current increases, and an opening of them if the 
current diminishes. To obtain this there is an apparatus acting 
as a sensitive ammeter. It consists of a relay acted on by the 
main current, the movable tongue of this relay, controlled by a 
spring, being in equilibrium with the normal cun-ent. If the 
intensity of the main current varies above or below the normal 
the relay tongue is attracted in one direction or the other. 
According to the direction in which the relay tongue moves 
current is sent through one or other of the armature windings 
of the 1 h.p. motor, and the Piccard relays open or close the 
turbine sluices. As the regulation acts on the four turbines 
simultaneously, a ver\^ small movement compensates for large 
variations of current. 

On account of the lightness of the moving parts of the 
turbines and dynamos, the action is in effect very like automatic 
regulation. When the resistance of the circuit is increased by 
the addition of motors the cuiTent diminishes, the torque also 
diminishes, and the speed of the turbines increases, thus tending 
to bring back the current to its normal value. The reverse 
effect occurs if the current increases. The motor and relay 
regulators have thus only to correct small variations. 

The Traiismittinfj Arrangements. — The current is transmitted 
by bare copper wires three-eighths of an inch in diameter, 
carried on posts, with porcelain and oil insulators. The 
greatest distance of transmission is thirty miles. Each circuit has 
a resistance of about 500 volts, so that the line eflBciency at full 
load is about 90 per cent. 

The Motors. — The motors are all Thury motors, and are from 
5 to 60 h.p. From 5 to 18 h.p. they are bipolar. The^r 
regulation is effected by shunting more or less of the current, 
and they are all placed in series. Larger motors are multipolar, 
and these are regulated by displacing more or less the points at 
which the current enters and leaves the motor. This has the 
effect of reducing the magnetic field, by causing some of the 
convolutions to be traversed in a direction reverse to the 


normal. Each motor is governed by a relay, and has a fly- 
wheel. A lever and counterweight is provided to adjast the 

The ratio of the effective work at the motors to the effective 
work of the turbines is stated to be 72 per cent. 

Transmismmi at BiberhsU near Solenre. — Messrs. Cuenod 
Sautter & Go. have also carried out a power transmission 
between Ronchatel and Biberist. Turbines of 360 h.p. have 
been erected, and the power is transmitted 28 kilometres to a 
paper mill. The generating station has two Thury continuous 
current dynamos, coupled to the turbines by Raffard couplings, 
and running at 275 revs, per minute. Each dynamo gives a 
current of 3,300 volts, and the two dynamos are coupled in 
series. The current is transmitted by a bare copper wire, | inch 
in diameter, on simple porcelain Jinsulators. The receiving 
station has two dynamos similar to the generators, making 200 
revs, per minute, and driving directly ^by Kaffatd couplings the 
machinery of the mill. An efficiency of 70 per cent, is 


Tlie Lauffeii-Ileilbronn Transmismon. — This transmission is 
interesting as an example of the 'Drehstrom' or three-phase 
method of transmission, and because an effort has been made to 
combine power supply with light supply. The owners of the 
Wiirtemburg Cement Works at Lauffen, on the Neckar, having 
surplus water power, conceived, in 1889, the idea of utilising it 
to supply electricity to Heilbronn, some six miles distant. They 
accepted plans for using the three-phase current with step-np 
and step-down transformei-s. ITie generating dynamo at Lanffen 
gives 4,000 amperes at 50 volts. This is transformed to 5,000 
volts for transmission, and then reduced to 1,500 volts in 
Heilbronn for distribution. Part is further reduced by second- 
ary transformers to 100 volts in a network for lighting. The 
charge for current used for lighting is 9d. per kilowatt hour. 
Current used for power purposes is charged at 4d. per kilowatt 
hour, or about 42/. per effective h.p. year of 3,000 working hoars. 
In November 1892, there were on the system only 11 motors, 
aggregating 32 h.p., and it is believed that the use of electricity 


for power purposes is not even now very considerable. To 
prevent sudden fluctuation of the lighting circuits due to 
switching in or out of the motors, fluid resistances are arranged 
so that full contact cannot be made in less than 1 5 to 20 seconds. 
Motors over 3 h.p. are coupled direct to the high pressure mains. 
It is curious that as the energy supplied by the water is costless, 
it has been found economical to use surplus current in times of 
light load for drying the clay used in the Cement Works. 

The New Electrical ScJieme at Geneva. — The hydraulic in- 
stallation at Geneva has already been described. The total 
amount of power available at Geneva will shortly be completely 
utilised. To obtain a further supply of power, a site has been 
found at 6 kilometres below Geneva (8 kilometres by the river) 
where a fall of 16 feet is available in high water, and 27 feet 
in low water, of the Rhone. The total amount of power avail- 
able is 14,000 h.p. It is proposed to utilise the power at this 
site by 15 turbines of 800 h.p. each. The works are already 
in progress. For transmission the electrical method is to be 
used, and there is no doubt that alternating currents will be 

Transmission of Power at the Gold King Mine, Teliwidsy 
CoUn-ado, — This installation is interesting as one of the first 
applications of alternating curi'ents to the transmission of power.* 
A Pelton wheel on a fall of 320 feet drives an alternate current 
generator at 833 revolutions per minute, giving 166 alternations 
per second. The current is carried by bare conductors on posts 
to a crushing mill three miles distant, where it drives a 
synchronizing motor of 100 h.p. at 3,000 volts. The generator 
has a composite field winding. Part of the field magnets are 
excited by direct current from a separate exciter. The rest 
are excited by a current from the armature, proportional to the 
main current, and commutated by the equivalent of a two-part 
commutator. The pressure at the terminals of the generator 
rises as the current increases, so as to compensate for the 
increase of line loss, and to keep the pressure at the motor 
constant. Generally no adjustment is required after the motor 
is started. The motor is similar to the generator, but the field 
current is obtained from a second winding parallel with the 

' See *LoDg Distance Transmission for Lighting and Power,* by C. P. Scott, 
Tram. Am. Soc. of Electrical Engineen, 1892. 


main coils on the armature. This current is coramutafced and 
passes through the field coils. To start the synchronizing 
motor a Tesla rotating field motor is used. This motor comes 
quickly to its normal speed, and then is used to bring the large 
motor up to speed. Both motors are belted to a countershaft, 
so that at the normal speed of the Tesla motor the synchronous 
motor is running above its normal speed. The large motor 
field magnets are then excited, and it runs as a self-exciting 
generator. The Tesla motor is switched oflF and the speed of 
the synchronous motor begins to fall off. At the moment when 
it has the same speed and phase relation as the distant generator 
(as shown by lamps), the main current is switched on. The 
small motor is disengaged by a friction clutch, and the load put 
on the synchronous motor by another clutch. Full load may be 
thrown on suddenly. The plant was started in June 1891. 
The aggregate time lost in the first nine months' working, from 
mishaps, was 48 hours. The plant has worked so well that it is 
to be increased by erecting a 750 h.p. generator and other 




Few persons can have seen Niagara Falls without reflecting on 
the enormous energy which is there continuously expended, or 
without some trace of a feeling of regret that a supply of motive 
power of such enormous commercial importance should be 
wasted. To any engineer it must have occurred that the con- 
stancy of volume of flow and small variation of levels, the height 
of the fall, the suitability of the rocks for engineering operations, 
the proximity of railways and access by water to the great lakes, 
all marked out Niagara as an ideally perfect site for a maim- 
facturing centre with factories worked by water power. 

A vast system of inland lakes extends halfway across the 
continent of North America, on the boundary between Canada 
and the United States. The lakes form storage reservoirs for 
the rainfall on an area of 2 tO,000 sq. miles, and they discharge 
at an almost uniform rate the collected drainage through the St. 
Lawrence to the Atlantic Ocean. Between Lakes Erie and 
Ontario, the last of the chain, the water flows through the 
Niagara River, which falls 326 feet in a distance of 36 miles. 
The average discharge through this short river has been 
estimated at 275,000 to '300,000 cubic feet per second. Such 
a volume of flow on such a fall would yield, if it could all he 
utilised, seven million horses power. 

TJie Earlier Works for Vtili sing Niagara, — From a very early 
time water mills were built in the rapids above the falls. 
These older mills, which greatly disfigured the appearance of the 
falls, have now been cleared away. In 1853, a more important 
scheme was started, and by 1861 the so-called hydraulic canal 
had been constructed. This canal, originally 35 feet wide and 
8 feet deep, receives water from the upper river at Port Day, 


nearly a mile above the falls. It conveys the water to a line of 
mills ranged along the blaff above the lower river. Tui^bines 
in these mills discharge their tail-water on the face of the cliff 
below the suspension bridge. By means of tliis old canal about 
6,000 h.p. has been utilised, during the last quarter of a century. 
The mills have been prosperous, and have been an im]X)rtant 
factor in the industrial life of the district. They employ more 
than a thousand operatives, io whom 70,000Z. a year is paid in 
wages. Recently, stimulated by the progress of another much 
larger scheme, the old hydraulic canal has been enlarged, and is 
now capable, it is said, of supplying turbines of 30,000 h.p. 

The Origin of the New Scheme. — About eight or ten years 
ago, the governments of the Province of Ontario and the State 
of New York began to be ashamed of the disfigurement of the 
falls, partly by mills and still more by ugly buildings erected by 
enterprising tradesmen. To restore and preserve the natural 
beauty of the falls, land was purchased and formed into State 
Reservations, within which objectionable buildings and other 
disfigurements were removed. Then arose a feeling that no 
more power could be obtained in the old way by mills along the 
bluff. The idea of a better method of utilising the falls is due 
to Mr. Thomas Evershed. He proposed to construct canals on 
unoccupied land a mile and a half above the falls. From these 
canals the water was to be supplied to turbines in vertical pits. 
The tail-water from the wheel pits was to be collected by under- 
ground tunnels into a great main tail-race tunnel, passing under 
the town of Niagara Falls and discharging into the lower river. 

In 1886, the Niagara Falls Power Company was fonned to 
carry out Mr. Evershed's plans, and a charter was obtained jfrom 
the State of New York, giving full rights to construct the 
works necessary to utilise 200,000 h.p. and to convey the power 
through any civil division of the State. A tract of 1,500 acres 
of land was acquired, about H miles above the falls. A subordi- 
nate company, the Cataract Construction Company, was con- 
stituted to caiTy out the necessary engineering works, and under 
its direction a great tail-race tunnel, 7,000 feet in length and 
capable of discharging the tail- water of turbines of 100,000 h.p., 
was commenced and has since been completed. The company 
have the necessary powers to construct a second tunnel whenever 
it is required. 


When Mr. Evershed's plan came to be considered in detail, 
Tit appeared that the removal of the mill sites up-stream and the 
discharge of the tail-water underground solved only part of the 
problem. Mr. Evershed contemplated only the supply of water 
to the mills and its removal afterwards, as in some other 
American water-power works. The mill-owners were to sink 
wheel-pits and erect the turbines. In part this plan is actually 
being carried out, but it is only suitable for factories requiring 
large amounts of power. The growth of an industrial city 
supplied with power in tliis way would necessarily be slow, and 
the capital sunk in the construction of the great tunnel and 
other works would for years yield no adequate return. The net- 
work of surface canals would be expensive and the network of 
underground tunnels still more costly. 

Obviously, it would economise the expenditure to develop a 
great part of the energy in power stations, by turbines of large 
size, aggregated in one place and placed under common manage- 
ment. It would facilitate the sale of the power, if manufacturers 
-could have it delivered to them in any quantity required without 
trouble or expense on their part. More important still, if once 
power distribution instead of water distribution was accomplished, 
there would be markets for the power in towns at a considerable 
•distance from Niagara. With so large an amount of power 
Available, it was of enormous importance to obtain access to 
markets where motive power could be disposed of. The more 
the problem of power transmission has been studied, the more 
probable it has become that a great part of the power developed 
at Niagara will be used in towns at a considerable distance. 
Provisional arrangements have already been made, with a view 
of distributing power from Niagara in Albany, Rochester, 
Syracuse, and Schenectady, and for working the navigation of 
the Erie Canal, provided present expectations as to the com- 
mercial possibility of transmission to great distances are proved 
to be well founded. 

IVie International Nicujara Commis.non. — In 1890, Mr. E. D. 
Adams, the President of the Cataract Construction Company, and 
Dr. Coleman Sellers visited Europe. It was during this visit, 
And chiefly as a result of examining the works for distributing 
power in Switzerland, that the importance of the problem of 
distributing power instead of water was first clearly realised. It 


was resolved to invite selected engineering firms to send in com- 
pletely worked out projects for the development and the trans- 
mission of power at Niagara, with estimates of cost. To secure 
a perfectly unbiassed consideration of these projects, a commission 
was formed consisting of Lord Kelvin, Professor E. Mascart of 
Paris, Colonel Turrettini of Geneva, and Dr. Coleman Sellers of 
Philadelphia. A sum of 4,500Z. was placed in the hands of the 
Commission, to be awarded, partly in preminms to all competitors 
who s«3nt in projects complying with the conditions, partly in 
prizes to those projects judged to be meritorious. Plans and 
instructions were prepared for the information of competitors. 
In the event of any project being adopted, the conditions of pay- 
ment were specified, both in the event of the machinery being 
constructed abroad and in the event of its being constructed in 
America under the superintendence of the competitor. A num- 
ber of projects were received, some for the utilisation of the water 
only, some for the transmission and distribution of the power^ 
and some both for the hydraulic and transmitting machinery. 
The more important and fully considered projects were fourteen 
in number ; and amongst these, in one distribution by wire ropes 
was proposed, in two distribution by compressed air. and in one 
distribution by pressure water. In all the other important pro- 
jects electrical distribution was assumed ; Messrs. Cuenod Sautter 
& Co., Messrs. Vigreux and Levy, and Messrs. Hillairet and 
Bouvier proposed continuous current working at constant pres- 
sure ; Messrs. Siemens proposed continuous current series work- 
ing at constant current and varjing potential ; Messrs. Ganz and 
Professor 6. Forbes proposed alternate current working. The- 
electrical plans of Messrs. Ganz and Messrs. Siemens were not 
worked out as detailed projects. The Oerlikon Company were- 
to have sent in plans for alternate current distribution, to accom- 
pany the hydraulic plans of Messrs. Escher Wyss& Co., of Zurich ; 
but they were not ready, and withdrew at the last moment. 

Touching as briefly as possible on matters now chiefly of 
historical interest, it may be stated that many of the competitors 
appeared before the Commission and explained and discussed 
their plans. It was assumed at that time that the greater part 
of the power would be distributed within two miles of the tur- 
bines, but that 25,000 or 50,000 h.p. might probably be sent to 
Buffalo, eighteen miles distant. This should be remembered in 


considering the conclusions of the Commission. The project of 
Professor Riedler showed that it was probable that power could 
be transmitted to Buffalo by compressed air, at a cost less than 
that of steam power produced in Buffalo. But the Commission 
decided in favour of transmission by continuous current electri- 
cal methods. They were of opinion that no one of the projects 
submitted could be carried out without modification. Amongst 
questions arising out of the plans submitted they decided in 
favour of placing the electric machinery above ground, and not 
in underground rock chambers. They also suggested that to 
secure safety and freedom from accident, the electric conductors 
should be placed in covered subways. 

The project at Niagara involves special difficulty in two ways. 
The works are of unj)recedented magnitude, not only in the 
aggregate, but in detail and in the size of individual units. 
Also the commercial condition is an absolutely governing one. 
Steam power is cheap in the district, and unless the water power 
can be distributed at a price less than that of steam power, the 
market for it would be very restricted. The fall is so advantageous 
that the cost of power at the turbine shaft will be extremely 
low. But if heavy charges are added to this due to the expendi- 
ture on distributing machinery, its development will be use- 
less. Many plans which have been proposed are impracticable, uot 
because they would fail to work mechanically and electrically, but 
because they are too costly. 

Tlie Scheme lunv being carried oat on the United States side, — 
The works now being carried out, and in part completed, com- 
prise — 

(1) The construction of a head-race canal 200 feet wide, 
1,500 feet long, and 17 feet deep. This is completed. 

(2) The construction of the great tunnel tail-race 6,700 feet 
long, 21 feet high, 18 feet 10 inches wide, with a slope of from 
4 to 7 feet per thousand. The net section of ^the tunnel is 386 
square feet, and it will discharge the tail-water of turbines de- 
veloping 100,000 h.p. The velocity of the water in the tunnel 
when in full operation will be 18 to 25 feet per second. The 
floor of the tunnel, at the point of discharge, is 205 feet below the 
cill of the entrance sluices at the upper river, giving an effective 
fall of 1 iO feet at the turbines. It was hoped that the rock was 
so strong that lining would be unnecessary. This proved not to 


be the case, and the tunnel is lined with four rings of brick in 
cement. For 200 feet back from the discharge portal it is lined 
with steel plates backed with concrete. Taking the discharge 
at 10,000 cubic feet per second, it amounts to about 3^ per 
cent, of the flow of the Niagara River. 

(3) Works for. the distribution of water to large consumers 
who erect their own machinery, with lateral tunnels to discharge 
the water into the tail-race tunnel, have been commenced. The 
first enterprise which took advantage of this arrangement was 
the Niagara Falls Paper Company (fig. 80). This company con- 
tracted to take 3,300 h.p. at once, with a right to take as much 
more subsequently. The mill has been built, and is said to 
be the largest paper mill in the world. The wheel-pit has been 
sunk 167 feet deep, and 28 feet by 43 feet in section. A verti- 
cal supply pipe, 13^ feet in diameter, conveys water to three 
Geyelin, inverted Jonval turbines of 1,100 h.p. each. The tail- 
water flows away, by a tunnel 7 feet diameter and 087 feet long, 
to the main tunnel. The power of the turbines is transmitted 
to the ground level by 10-inch vertical shafts ; these drive the 
mill shafting by bevil wheels, which have 20 inches width of face 
and 6 inches pitch. The upward pressure of the water on the 
turbine wheels balances the weight of the vertical shafts. The 
shafts run at 260 revolutions per minute. The three turbines 
were recently started with success, and the mill is now in 

The Paper Company get their land and water power 
(exclusive of the cost of the machinery) at ,^8, or about Sbs. per 
h.p. per annum, having the right to use the power 21 hours per day. 
The Pittsburgh Reduction Company, who control patents for 
the production of aluminiiun electrically, have acquired a site 
for works and the right to use 6,000 h.p. Other undertakings 
also are negotiating for sites and water supply under similar 
conditions. These works are carried out on the plan originally 
intended by Mr. Evershed. 

(4) A large power station has been commenced, from which 
energy will be transmitted electrically both to consumers at 
Niagara and to Buflialo, and perhaps to towns more distant. 
The power house, shown in fig. 80, is placed alongside the head 
race canal, and is designed to contain 10 turbines of 5,000 h.p. 
each. Three of these have already been constructed. A wheel- 



pit or slot has been sunk, which at present is 140 feet in length, 
21 feet in width, and 178 feet in depth, having room for four 
turbines. It will be extended in length as required. Over 
this the first section of the power house has been erected, and 
it contains a 50-ton electric travelling crane, commanding the 
whole floor, which is being used in erecting the .machinery. 

The 5,000 h.j). Tarbhies. — Soon after the meeting of the 
Commission it was decided that the first turbines for the power 

Fig. 80. 

station should be of 5,000 h.p., running at 250 revolutions per 
minute, to suit the requirements of the electrical engineers. It 
was decided to place the turbines at the bottom of an open slot 
or wheel-pit, 175 feet in depth, and transmit the power to the 
dynamos at the ground level by vertical shafts. Designs by 
Messrs. Faesch & Piccard, who conjointly with Messrs. Cuenod 
Sautter & Co. sent in the combined project which received the 
highest award of the Commission, were selected. The first three 
turbines have been constructed by Messrs. J. P. Morris & Co., of 



Fig. 81. 

Philadelphia, under 
the superintendence of 
Messrs. Faesch & Pic- 

fi^ ;• card, and they are now- 

being erected. 

Hasty critics have 
assumed that, whatever 
difficulty there may be 
about the transmission 
of the power, the hy- 
draulic part of the pro- 
blem is of a very ordi- 
nary character. That 
is not so, and the way 
in which the essential 
conditions have been 
met, without sacrifice 
of efficiency and by 
arrangements of very 
great simplicity, re- 
flects great credit on 
the mechanical skill 
and judgment of 
Messrs. Faesch & Pic- 
card. The water de- 
scends the wheel-pit 
to each turbine, fig. 8 J , 
in a supply pipe 7 ft. 
Gins. in diameter. The 
supply pipe bends at 
right angles at the 
bottom and delivers 
the water between a 
pair of twin outward 
flow turbines. The hy- 
draulic pressure on the 
top cover of the wheel 
chamber, amounting 
to GO or 70 tons, is 
used to support the 


weight of the vertical shaft; and the revolving field magnets of the 
dynamos, while the pressure on the bottom of the wheel chamber 
is supported by the foundations. No mechanical arrangement 
of footstep which could have been designed would have sup- 
ported so great a weight, on a shaft running at 250 revolutions. 
One considerable diflSculty of the mechanical problem is thus 
met by means involving absolutely no expense, and perfectly 
permanent in action. The adjustment of the exact amount of 
unbalanced upward force can be left to be decided at the last 
moment. Collar bearings on the shaft support the small 
difference of load and upward thrust which may unavoidably 

The turbine wheels are of bronze, 6 ft. 2 ins. in diameter, 
and by adopting twin turbines the whole arrangement is made 
compact, and the double condition of supplying the necessary 
power at the required speed has been met, without sacrifice of the 
hydraulic requirements on which the efficiency of the turbines 
depends. Efficiency at * part gate ' is not important at Niagara, 
and two ring sluices, on the outside of the turbine wheels, have 
been adopted for regulating the power. The pressures on these 
being balanced, they are without difficulty put »nder the control 
of a relay speed governor, notwithstanding their great size. To 
give the governor time to act on the sluices, a fly-wheel of con- 
siderable power is necessary. This fly-wheel has been obtained 
by making the field magnets of the dynamos the revolving part, 
the comparatively light armature being stationary. The speed 
regulation of turbines driving diynamos is of the greatest im- 
portance, especially as alternating dynamos in parallel are to be 
adopted. Experience has shown that the speed regulation of 
turbines of large size by governors presents a mechanical 
problem of considerable difficulty. The speed governor is a 
relay governor of a type used by Messrs. Faeseb & Piccard in 
other installations on a smaller scale. They have so much 
confidence in its action that they have given a guarantee that 
when a quarter of the load, that is 1,250 h.p., is suddenly thrown 
off, the speed variation will not exceed 2 per cent. 

The relay is a mechanical relay. A shaft geared to the 
turbine sluices carries two pulleys, oi'dinarily running loose and 
driven by an open and crossed belt. Either pulley can be 
geared instantly to the sluice shaft by a friction brake, which 



holds one of a train of three bevil wheels. The brakes are pnt 
in action by a sensitive pendulum governor, acting on ratchet 
gear. The governor merely puts the ratchet click in gear. 
Hand regulation is provided. 

The EUclrical Installation.^ — At the time of the meeting of 
the Niagara Commission, the continuous current method of 
working was the only one which had been practically used in 
any large distribution of motive power. The gradual growth, 
subsequently, of a conviction that distant transmission was of 
the greatest importance at Niagara led to a decision, in 1893, to 
adopt an alternating current system. For the Niagara work, 
continuous current methods involved too great a limitation of 
the electrical pressure for economical distant transmission. The 
alternate current method has one special advantage, in the 
possibility of varying the electrical pressure by statical trans- 
formers. Initially, the adoption of an alternate current method 
was open to the objections that it involves greater stress on the 
insulation, and that motors for alternate currents were not 
altogether satisfactory for general use, and were almost untried. 
Progress in the construction of such motors has been made in 
the last year or two, and the objection now has less force. Some 
doubt also existed at first as to the satisfactory working of 
alternate current generators in parallel. It would be extremely 
inconvenient if the generators could not be used in groups 
according to the demand for current. Later experience shows 
that alternate current generators can be constructed to work 
synchronously, in parallel, in a perfectly satisfactoiy way. The 
sDccess of the three-phase motor at Frankfort, and of the two- 
phase motors of Tesla in America, has led to the selection of a 
two-phase system of working. Such a system with two in- 
dependent circuits is equally available for working two-phase 
and one-phase motors, and for the distribution of part of the 
current at any point for ordinary alternate current lighting. 
There remain certain applications, such as tramway working, 
where continuous currents are most suitable. It appears 
probable that methods will be perfected for commutating 
alternate currents after their transmission and transformation to 

' See, for much fuUer details, the paper by Piofesaor Q. Forbes, * Electrical 
Transmission of Power from Niagara Falls,* Journal Inst, of Eleetrieal 
Engineeri, toI. zxii. 


low tension. In any case, motor transformers can be used to 
obtain continuous currents when necessary. 

Acting on the advice of their electrical adviser, Professor G. 
Forbes, a departure is to be made from previous practice in the 
electrical system by reducing the frequency of the alternations.' 
Hitherto, alternating currents have been used almost exclusively 
for lighting, and high frequency (100 to 133 periods) has been 
commonly adopted, partly to avoid flicker in the lamps, partly 
because it reduces the cost of transformers. Messrs. Ganz have 
for some time adopted 42 periods per second. Professor Forbes 
proposed to use 16f periods per second; but, after negotiation 
with the constructors of the dynamos, a frequency of 25 periods 
has been selected. Professor Forbes believes that low frequency 
will make parallel working of the dynamos more easy, and 
improve the efficiency and facilitate the construction of motors. 
It will reduce the loss due to the unequal distribution of the 
current in the section of the transmitting conductors, a matter 
of importance with the large conductors required at Niagara. 

Professor Forbes is prepared to construct dynamos to work 
up to 20,000 volts, in which case step-up transformers would be 
unnecessary for the transmission to Buffalo. But, proceeding 
more cautiously, it has been decided that the three first 5,000 h.p. 
dynamos, now being constructed by the Westinghouse Company, 
shall work at 2,000 volts, and deliver a two-phaee current. For 
the district at Niagara, electricity will be distributed at 2,000 
volts; for transmission to Buffalo it will be transformed to 
10,000 volts. 

A subway large enough to walk through has been con- 
structed for a third of a mile through the land of the Company. 
Beyond that the transmission to Buffalo will be by bare copper 
conductors on insulators carried on posts. 

The Scheme on the Canadian side, — ^The idea of utilising 
power on the Canadian side was fii*st discussed some years ago. 
Plans were prepared in which Pelton water-wheels driving 
dynamos were to be placed in rock chambers underneath the 
river, just above the Horseshoe Fall, and the tail- water discharged 
by tunnels, directly on the face of the escarpment forming that 
fall. Various vague statements were made as to transmitting 

> Mr. Tesia is also an advocate of low frequency for motor work. See 
Inveatums, Besearehet, and Writings of Nikola Testa ; New York, 1894 ; p. 8. 


the electricity to Toronto and Montreal at enormous electric 

Recently, by charter from the Canadian Government, rights 
have been acquired to develop 250,000 h.p. on the Canadian 
side, by a company acting in association with the company 
developing power on the other aide. The Cataract Construction 
Company undertakes the engineering work, and is under con- 
tract to complete a considerable power development within three 
years. The power station will be within the Victoria Park 
Reservation, and as at present planned will be similar to the 
power station on the other side. The conditions for developing 
the power on the Canadian side are more favourable than on the 
American side. Deep water in the river comes close to the 
shore, and the head-race and tail-race tunnels will be much 
shorter. Two tunnels are contemplated, each large enough for 
125,000 h.p., and the first of these does not need to be more 
than 400 feet in length. 

[Note. — Since this chapter wa« in type a paper on the cost of power at 
Niagara, by Messrs. £. J. Honston and A. E. Kennelly, has been received. 
Allowijig 2.0 per oent. interest and depreciation on the hydranUc works and 
turbines, they estimate the cost of power at Niagara at 7<. per effective h.p. 
per anniun« if working constantly at fall load ; or 12f . per h.p. per annum if the 
average load is 60 per cent, of the maximum. Then, assuming a three-phase 
transmission at a maximum pressure of 50,000 .volts, and allowing interest on 
dynamos, transformers, motors, and line, and also for loss in transmission, 
they calculate the cost of the power delivered at various places. In all cases 
the average load is taken at 60 per cent, of the maximum. At Buffalo, 15 
miles distant, they estimate the power delivered will cost 21. 13f . per h.p. per 
annum. At Albany, 330 miles distant, they estimate the eost at -5^. lOf. per 
h.p. per annum.} 


Abraham's air meter, 194 

Accumulator, 69 ; electric, 71 ; hy- 
draulic, 91, 130, 142 

Air compressors, 164, 171 ; water injec- 
tioD, 177; compound, 179; action 
in, 223 

Air mains, losses in transmission, 184 ; 
forms of, 186 ; cost of, 1 87 ; joints 
for, 187; loss of pressure in, 211; 
in Paris, 201 ; flow of air in, 212 ; 
experiments on Paris, 219 ; calcula- 
tion of, 237 

Air motors, 188, 202, 232 ; at Serain^c, 
1 96 ; action of, 232 ; experiments on, 

Alternating currents, 268, 306 

Alternating carrent systems, 283, 294 

Antwerp, Compagnie hydro- 61ectrique, 

Back pressure in steam engines, 41 

Battery storage, 71 

Bellegarde, 82, 113, 124 

Berne compressed air tramways, 204 

Biberist electric transmission, 294 

Birmingham, coal consumption in 
small engines, 28 ; hydraulic system, 
146; air mains, 187; compressed 
air system, 166 ; compressors, 227 

Boilers, evaporatiye power, 21 ; waste 
in irregular working, 52 

Boudenoot vacuum system, 164, 199 

Bradford electric station, 65 

Bramwell, Sir F., 19 

Calorific value of fuels, 10 ; of ash- 
bin refuse, 16 ; of gas, 259 
Canadian scheme, 307 
Carels Fr^res, Antwerp, engines, 102 
Carnegie on natural gas, 267 
Castermans on Herstal factory, 287 
Cederblom, Prof., on steam turbine, 57 

Central stations, advantages of, 4 
Coal, calorific value of, 10 
Coal consumption in engines, 23 ; in 
lighting stations, 27; in small 
engines, 28 
Colladon compressor, 156, 164, 178, 

Compressed air, 163 ; history of, 163 
Compressed air storage, 69, 197 
Compression, isothermal, 173; adia- 

batic, 174 
Compressors, 177; compound, 179, 

228 ; at Paris, 203 ; action of, 223 
Condensation in cylinders, 29 
Conductors of electricity, 266 
Continuous current systems, 280, 287 
Comer on compressed air, 70, 197 
Cost of power at Geneva, 6 ; of liquid 
fuels, 11 ; of working destructors, 
15 ; of steam power, 58 ; of engines 
and boilers, 68 ; of gas engines, 60 ; 
of steam power in central stations, 
61 ; of accumulator batteries, 71 ; 
of thermal storage, 74; of water 
and steam power, 89; of water 
power at Geneva, 90 ; of power at 
Bchaffhausen, 121 ; of pressure 
water, 140, 143; of power on hy- 
draulic system, 144, 146 ; of power 
at Zurich, 149 ; of power at Geneva, 
169; of air mains, 187; of power 
from compressed air, 194 ; of steam, 
247 ; of gas, 256 ; of electric con- 
ductors, 270; of electric trans- 
missions, 273 ; of power at Niagara, 
Cotterill, Prof., on cylinder conden- 
sation, 44 
Creusot, engine experiments, 41 
Crompton on coal consumption in 
lighting stations, 27 ; on IoeuI curves, 
Cylinder waUs, action of, 29, 44 



Dbssau electric station, 256 

Destructors, 13, 78 

Distribution of power by compressed 

air, 163; theory of , 211; practical 

calculations for, 237 
Distribution of power by steam, 245 ; 

by gas, 255 ; by electricity, 263 
Dowson gas, 13, 258, 261 
Dwelshauvers Dery, on action of 

cylinder waUs, 31 

Efficiency, mechanical, of engines, 
39; thermal, of engines, 43; of 
turbines, 102, 104 ; of cable trans- 
mission, 114; of hydraulic trans- 
mission, 136; of air compressors 
and motors, 176, 223, 231 ; of com- 
pressed air transmission, 185; of 
electric transmission, 269 ; of Her- 
stal installation, 288; of vacuum 
system, 199 ; of Shone system, 209; 
of Genoa installation, 294 
Ejectors for sewage, 209 
Electric central stations, coal con- 
sumption in, 28 ; load curves of, 38 
Electric motors, 283, 293, 306 
Electric station at Ziirich, 149; at 
Geneva, 158; at Genoa, 290, 296; 
at Heilbronn, 294; at Telluride, 
Ellington, pipe joints, 134 ; efficiency 

of hydraulic system, 137 
Emerson on testing turbines, 87 
Emery, Dr., on cost of water power, 
81, 89; on steam distribution, 246 
Engine efficiency, 23, 29 ; friction, 40 
Evaporative power of boilers, 19 
Evershed on Niagara scheme, 298 
Ewing, Prof., on steam turbine, 56 

Fabsch and Ficcabd governor for 

turbines, 107, 293, 305 
Filters, 143 
Flow of water in pipes, 133 ; of air in 

pipes, 212 ; of gas in pipes, 259 ; of 

electricity, 268 
Flywheel storage, 68 
Forbes, Prof. G., on destructors, 18 ; 

on utilisation of Niagara, 306 
Francis, J. B., on water power, 85 
Frankfort-Lauffen experiment, 3, 264 
Fribourg rope transmisAon, 122 
Fryer's destructor, 14 
Fuel, solid, 9 ; liquid, 11 ; gaseous, 12, 

259 ; town refuse, 13 

Ga8 for power purposes, 12, 255 

Gas engines, 13,30, 54, 94, 146, 193, 257 

Gasholder storage, 68 

Gas Light and Coke Company, load 

curve, 37 
Gas motor. See Gas Engine 
Geneva power distribution, 6, 82, 93, 

149 ; electric installation, 295 
Genoa electric transmission, 290 
Girard turbines, 103 ; pumps, 148, 154, 

Gokak rope transmission, 124 
Governors for turbines, 105, 158, 292, 

Greathead injector hydrant, 141 
Greenside mines electric transmission, 

Gutermuth, Prof., experiments on re- 
heaters, 193; on Paris air mains, 
217; on compressors, 231; on air 
motors, 235 

Halpin, thermal storage, 72 

Hanarte on compressed air, 165; 
compressor, 178 

Heat exchange in cylinder, 30 

Heating power of fuel, 9; of town 
refuse, 16 ; of gas, 259 

Heilbronn transmission, 285, 294 

Herschel, C, on Holyoke water power, 

Herstal small-arms factory, 287 

Hillairet, Dom^ne installation, 264 

Him on superheating, 32 

Him, C. F., on telodynamic trans- 
mission, 108 

Holden, use of liquid fuel, 11 

Holly, steam distribution, 245 

Holyoke, water power at, 85 ; testing 
flume at, 87 ; tests, 103 

Horsfairs destructor, 15 

Hughes and Lancaster, tramway 
system, 206 

Hull hydraulic system, 139 

Hydraulic motors, 95 

Hydraulic Power Company, 5; coal 
consumption of engines, 27; load 
curves, 36; accumulator, 92; de- 
scription of works of the, 140; 
comparukon of hydraulic and air 
systems, 185, 197 

Hydraulic storage, 69 

Hydraulic transmission, 129; high 
and low pressure systems, 131 ; 
efficiency of, 136 ; general arrange- 
ment of, 138; examples of high 
pressure systems, 139 ; examples of 
low pressure systems, 146 

Indicated effective horse-power, 39 
Internal furnace engines, 54 



Jacket, Bteam, 26, 32 
Jackets for compressor?, 177, 227 
Joint for air mains, 187 ; for hydraulic 
mains, 135 

Eapp, G., on rope transmission, 122 ; 

on electrical transmission, 263 
Keep on refuse destructors, 17 
Kelvin, Lord, economical section of 

conductors, 271 
Ecnnedj, J., origin of factory system, 

Kennedy, Prof., experiments on boilers, 

Kraft, J., use of compressed air, 196 

Laval steam turbine, 56 
Lawrence, water power at, 85 
Leavitt, water piston compressors, 178 
Liverpool hydraulic system, 146 
Load curve and load factor, 35 
London hydraulic system, 5, 137, 140 

Mains, for water, 133; for air, 184; 

experiments on, 217; for steam, 

249 ; for electricity, 278 
Mair-Bumley, engine trial, 30 
Manchester hydraulic system, 146 
Mannesman pipes, 135, 162 
Marseilles, efficiency of hydraulic 

plant at, 137 
Mekarski tramway system, 164, 205 
Meters for water, 141 ; for air, 167, 

194 ; for steam, 248 
Miller, Oskar von, 72 
Mines, compressed air in, 163, 197 
Moser, H., SchaShausen works, 116 
Motors for air, 188, 232 

Nevada electric transmission, 289 
New York steam system, 246 
Niagara, 82, 265 ; utilisation of, 297 
Niagara Commission, 265, 299 

Obeburskl rope transmission, 115 
Ochta rope transmission, 116 
Oyonaz electric installation, 266 

Pa BIS compressed air system, 164, 
185, 200; experiments on air mains 
in, 219 ; compressors, 227, 229 

Parry, J., on hydraulic power, 146 

Pardons, steam turbines, 55 

Pearsall, H. D., hydraulic compressor, 

Petit and Boudenoot vacuum system, 

164, 199 
Petroleum engines, 54 
Piccard, relay governors, 105, 158, 

293, 305 
Pipes for air, 185, 212; joints for, 

187; loss of pressure in, 222, 238 
Pipes for pressure- water, 131 ; friction 

in, 133; strength of, 134; joints 

for, 135; weight and cost, 136; 

steel, 136, 162 ; joints for, 135 
Pipes for steam, 249; variators for, 

Pneumatic transmission, 163 
Polyphase systems, 284 
Popp compressed air system, 164 
Portsmouth Dockyard compressed air 

system, 70, 197 
Preller on Berne tramways, 204 
Pressure engines, hydraulic, 96, 139, 

152, 158 
ProU, combination of gas and air 

motorii, 193 
Pulleys for rope transmission, 113 

Baffabd coupling, 160 

Beheating compressed air, 191, 233 

Belay, hydraulic, 157 

Beservoir storage, 92, 180, 149, 151, 

Beuleaux, Prof., rope transmission, 

Biedler, Prof., valves, 162, 183; com- 
pressors, 201, 229 ; experiments on 
Paris mains, 217; experiments on 
efficiency of compressors, 231 ; ex- 
periments on efficiency of motors, 

Bieter Brothers, brake for turbines, 
107; rope transmission, 110, 115, 
122,127; turbines, 291 

Bigg, A., hydraulic motor, 97 

Bope transmission, 108 ; at Oberursel, 
1 15 ; at Ochta, 116 ; at Schaffhausen, 
116; at Fribourg, 122; at Belle- 
garde, 124; at Gokak, 124; at 
Genoa, 291 

Botary motors at Paris, 190 

Bysselbergh, von, hydro-electric sys- 
tem at Antwerp, 161 

Sakkbt, Capt., steam consumption 

with variable load, 52 
Savage on Temi works, 165 
Schatfhausen, rope transmission at, 

81, 116; electric transmission, 122 
Bchmid motors, 96, 158 
Schwoerer, B., superheater, 33 
Secondary batteries, 71 



Shone, I., sewage system, 164, 207 
Siemens, A.,continuons and alternate 

current systems, 287; series system, 

Siemens, Dr. VV., on electric trans- 
mission, 263 
Soci6t6 Cockerill, use of compressed 

air, 195 
Solignac on compressed air, 168, 186 
Steam consumption in engines, 23; 

curves of, 46; with variable load, 

Steam distribution in towns, 245 
Stockalper, experiments on air mains, 

212, 220 
Storage of energy, 67, 90, 130 
Subways for electric mains, 280, 307 
Superheating, 26, 32 
Swain, G. 1?\, water power in America, 


Tellubidb electric installation, 265, 

Tclodynamic transmission, 108, 126 

Temi steel works, 165, 198 

Tesla motors, 283, 296 

Testing flume at Holyoke, 87 

Thermal storage, 72 ; cost of, 74 ; use 
with destructors, 78; use at gas- 
generating stations, 78 

Thomson, Prof. J., turbine, 103 

Thurston, Prof., 87 

Thwaite on gas motive-power stations, 

Town refuse as fuel, 13 

Tramways at Beme, 204; Hughes 
and Lancaster system, 206; electric, 

Trewby on gas consumption, 256 

Turbine, steam, 65 

Turbines, 99; at Schaflfhausen, 118; 
at Gokak, 126; at Zarich, 148; at 
Geneva, 154; at Nevada, 289; at 
Genoa, 291; at Telluride, 295 ; at 
Niagara, 303 

Turrettini, Th., Geneva system, 6, 
152; hydraulic relay, 157 

United States, water power of, 83 

Vacuum method of transmitting 

power, 164. 199 
Variatcrs for steam mains, 250 

Wateb power, 80 ; in United States, 
83; relative cost of water and 
steam power, 89; at Ziirich,'14f ; 
at Geneva, 149; at Antwerp," 160; 
at Schaffhausen, 116; at Fribonrg. 
122; at Bellegarde, 124; at Gokak, 
124; in Nevada, 280; at Gencka, 
290 ; in Colorado, 295 ; at Niagara^ 

Watson on refuse destructors, 14, 17 

Weissenbach, on water power, 81 

Willans engine, 48 

Willans' law of steam consumption in 
engines, 50 

Wind power, 8 

Witz, on gas engines, 12; on action 
of cylinder walls, 30 

Zeunsb, Prof., tests of turbines, 103 
Ziegler ropp transmission. 111, 114 
Zilrich, 82 ; power transmission at, 140